CN111183225A

CN111183225A - Improved AAV capsid production in insect cells

Info

Publication number: CN111183225A
Application number: CN201880058315.8A
Authority: CN
Inventors: D·J·F·杜普莱西斯; O·特布雷克; S·M·博斯马; H·P·A·彼得里; J·卢贝尔斯基
Original assignee: Uniqure IP BV
Current assignee: Uniqure IP BV
Priority date: 2017-07-20
Filing date: 2018-07-20
Publication date: 2020-05-19
Anticipated expiration: 2038-07-20
Also published as: AU2018303230A1; WO2019016349A1; EA202090332A1; KR20200045466A; EP3655538A1; MX2020000663A; US20200248206A1; CA3069369A1; US11667931B2; CN111183225B; JP2023093577A; ZA202000273B; JP2020532286A; US20240093231A1

Abstract

The present invention relates to the production of adeno-associated viral vectors in insect cells. The insect cell thus comprises a first nucleotide sequence encoding an adeno-associated virus (AAV) capsid protein, wherein the initiation codon for translation of the AAV VP1 capsid protein is AUG. A selective out-of-frame start codon is placed upstream of the VP1 open reading frame such that the initiation of translation of the VP1 protein is modulated, i.e., reduced, to allow production of VP1: VP2: VP3 in good stoichiometric ratios, resulting in high titers of AAV.

Description

Improved AAV capsid production in insect cells

Technical Field

The present invention relates to the production of adeno-associated virus in insect cells and to adeno-associated virus providing improved infectivity. The invention also relates to means and methods for libraries of adeno-associated viral vectors.

Background

Adeno-associated virus (AAV) can be considered one of the most promising viral vectors for human gene therapy. AAV has the ability to efficiently infect both dividing and non-dividing human cells, the AAV viral genome integrates into a single chromosomal site in the host cell genome, and most importantly, even though AAV is present in many humans, it has never been associated with any disease. In view of these advantages, recombinant adeno-associated virus (rAAV) is being evaluated in gene therapy clinical trials for hemophilia B, malignant melanoma, cystic fibrosis, and other diseases. A large number of clinical trials in Europe and approval of gene therapy drugs, such as Alispolene tipivovec (

uniQure), which is expected to make AAV a major component of clinical practice.

In general, there are two main types of production systems for recombinant AAV. On the one hand, there are conventional production systems in mammalian cell types (e.g.293 cells, COS cells, HeLa cells, KB cells) and on the other hand there are production systems using insect cells.

Mammalian production systems suffer from several disadvantages, which may include a limited number of rAAV particles produced per cell (10)⁴Of the order of magnitude (in Clark, 2002, Kidney Int.61(Suppl.1): review in 9-15) and cumbersome large-scale preparation. For clinical studies, more than 10 may be required¹⁵The rAAV particle of (a). To produce this number of rAAV particles, it would be necessary to use about 10¹¹Human 293 cells in culture, corresponding to 5000 cells at 175-cm²Cells in flasks, meaning transfection of up to 10¹¹And 293 cells. Thus, large scale production of rAAV using mammalian cell culture systems to obtain materials for clinical trials has proven problematic, with production at large commercial scaleIt may be less feasible. Furthermore, there is always the risk that vectors produced in mammalian cell culture for clinical use will be contaminated with undesired, potentially pathogenic substances present in mammalian host cells.

To overcome these problems of mammalian production systems, AAV production systems have been developed using insect cells (urabeetal, 2002, hum.131935-1943; US 20030148506 and US 20040197895). The AAV wild-type capsid from the wild-type virus consists of about 60 capsid proteins, VP1, VP2, and VP3 in a stoichiometric ratio of about 1:1: 10. Without being limited by theory, it is believed that the stoichiometric ratio is important for the recombinant AAV to obtain good titers (i.e., good transduction). In wild-type viruses, i.e. in mammalian cells, a stoichiometric ratio of about 1:1:10 of the three AAV capsid proteins (VP1, VP2 and VP3) was achieved, which depends on a combination of alternating use of two splice acceptor sites and less optimal use of the ACG start codon for VP 2. However, in order to produce AAV in insect cells, it is necessary to make modifications because the expression strategies found in mammalian cells do not replicate in insect cells. In order to obtain improved production of capsid proteins in insect cells Urabe et al (2002, supra) used a construct that was transcribed into a single polycistronic message that was capable of expressing all three VP proteins without splicing and in which the first translation initiation codon was replaced by the codon ACG. WO2007/046703 discloses further improvement of infectivity of rAAV vectors produced by baculoviruses (baculoviruses) by further optimizing the ratio of AAV capsid proteins in insect cells.

Urabe et al (J.Virol.,2006,80(4):1874-1885) reported that AAV5 particles produced in a baculovirus system using ACG as the initiation codon for VP1 capsid protein had poor transduction efficiency or potency, and that mutation of position +4 in the AAV 5VP1 coding sequence to a G residue did not increase infectivity, in contrast to AAV2 with VP1 expressed from the ACG initiation codon. Urabe et al constructed a chimeric AAV2/5VP1 protein in which the N-terminal part of at least 49 amino acids of AAV 5VP1 was replaced by the corresponding part of AAV2 VP1, which improved the transduction properties of the virion.

In yet another approach, expression of AAV capsid proteins is increased by inserting one or more amino acid residues in the AAV capsid coding sequence between a suboptimal (non-ATG) translation initiation codon and a codon encoding an amino acid residue corresponding to amino acid residue at position 2 of the wild-type capsid amino acid sequence (Lubelski et al, WO 2015137802).

Despite improvements in insect cell-based production of capsids for making AAV gene therapy vectors for use in medical therapy, there is a need to further improve AAV capsid production and to provide new methods to select improved AAV capsid constructs for expression in insect cells.

Disclosure of Invention

Brief description of the invention

The inventors surprisingly found that AAV capsids can be efficiently produced in insect cells from expression constructs encoding transcripts of VP1, VP2, and VP3 proteins from overlapping reading frames, wherein VP1 is translated from an AUG start codon. The prior art constructs containing the ATG initiation codon do not produce a VP1: VP2: VP3 ratio of about 1:1:10 as observed in wild type AAV, and therefore, without being limited by theory, do not produce a potent AAV. The expression constructs identified in the present invention allow for the efficient production of large numbers of high titer AAV gene therapy vectors for medical treatment in insect cells. Such vectors are at least similar if there is no improvement with respect to titer and quantity relative to AAV gene therapy vectors generated from alternative (alternative) initiation codons, such as CTG or GTG (see fig. 4).

Thus, the constructs of the invention contain an additional outer frame start codon 5' of the VP1ATG start codon, which clearly results in a reduction of the initiation of translation at the VP1 start codon, which allows for translation of sufficient amounts of VP1, VP2 and VP 3. Without being limited by theory, such constructs may allow for the expression of the VP1, VP2, and VP3 amino acid sequences, as they are found in wild-type viruses.

As shown in the examples, such constructs were identified by using a library of AAV capsid expression constructs for insect cells. The selection of constructs, which firstly require highly efficient production of AAV capsids in insect cells and secondly require high infectivity on selected target cells. Thus, the present inventors also provide highly efficient selection methods to provide AAV capsid expression constructs with improved properties (e.g., improved production and/or improved infectivity).

Accordingly, in a first aspect, the present invention provides a nucleic acid construct comprising an expression control sequence for expressing a nucleotide sequence comprising an open reading frame in an insect cell, wherein the open reading frame sequence encodes:

i) adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP 3; and

ii) the AUG translation initiation codon for VP 1;

wherein the nucleotide sequence comprises a selector start codon upstream of the open reading frame, the selector start codon being out-of-frame with respect to the open reading frame. In other words, the selective initiation codon is preferably 3N +1 or 3N +2 nucleotides upstream of the initiation codon.

In another aspect, the invention provides a method of providing a nucleic acid construct encoding a capsid protein of a parvovirus for production in an insect cell, said nucleic acid construct having one or more improved properties, said method comprising:

a) providing a plurality of nucleic acid constructs, each construct comprising:

a nucleotide sequence encoding a parvoviral capsid protein operably linked to an expression control sequence and at least one parvoviral Inverted Terminal Repeat (ITR) sequence flanking said nucleotide sequence encoding a parvoviral capsid protein operably linked to an expression control sequence;

b) transferring the plurality of nucleic acid constructs into an insect cell capable of expressing the Rep proteins of the parvovirus;

c) placing the insect cell under conditions that allow expression of the capsid protein of the parvovirus and the rep protein of the parvovirus such that the nucleic acid construct can be packaged in the capsid of the parvovirus to provide virions of the parvovirus;

d) recovering parvoviral virions from the insect cells and/or insect cell supernatants;

e) contacting a virion of the parvovirus with a target cell to allow infection of the target cell;

f) recovering the nucleic acid construct from the target cell.

Drawings

FIG. 1: schematic representation of the library generation and selection process. (a) First, a DNA library is provided. In this particular example, a library of expression constructs having multiple start codons (XXX) for AAV 5VP1 and random nucleotides (N) at selected positions (SEQ ID NO:71) is listed for examples of such constructs (1 is SEQ ID NO: 1; 2 is SEQ ID NO: 63; N is SEQ ID NO: 65); (b) the DNA library is transferred into a vector construct with an expression cassette with a promoter (P) for expression of the capsid proteins (Cap (VP123)) of AAV 5VP1, VP2 and VP3 flanked by two AAV Inverted Terminal Repeats (ITRs) to allow encapsidation in the AAV capsid. Also provided are expression cassettes for Rep52 and Rep 78; (c) the Cap and Rep constructs are then transferred to insect cells, in this case Sf9 cells. The transfer can be by a baculovirus vector, which allows control of multiplicity of infection; (d) thus, in insect cells, the expressed Rep52 and Rep78 proteins replicate and encapsidate the AAV vector genome containing the capsid expression cassette. As mentioned, when using baculovirus vectors, multiplicity of infection can be well controlled and for Cap constructs it is preferable to keep well below 1 to have on average only one library member per Sf9 cell to avoid cross-packaging. Only the Cap expression cassette that effectively produces the capsid will encapsidate the vector genome; (e) the capsids containing the vector genome are then tested for infectivity, i.e., efficient transfer of the vector genome to the target cells. Vector particles with a vector genome may for example be non-infectious, whereas vector particles with a vector genome and a VP1: VP2: VP3 ratio of about 1:1:10 are highly infectious. In this example, the HeLaRC32 cell line, which is also capable of replicating the AAV vector genome, was used. The vector genome can then be identified from the target cell. For example, the vector genomic sequence or a portion thereof containing the sequence of the changes indicated in (a) can be determined. Alternatively, the identifier sequence may be determined to identify a library member in (a) that has undergone successful infection of the target cell.

Thus, the combined steps (c) and (e) allow for the selection of capsid expression constructs allowing for efficient production in insect cells and the production of infectious virions in target cells. Selected expression constructs that predominate in the population may be particularly suitable candidates. The selected candidate expression constructs (without flanking ITRs) can then be used, for example, in a baculovirus vector or inserted into a cell line, to generate an AAV gene therapy vector.

FIG. 2: in these figures, the percentage of library members (y-axis) with a particular start codon (x-axis) at each stage of the selection process is shown; A) this figure shows the distribution of the start codons for a plasmid DNA library made from expression cassettes flanked by AAV ITRs. Prevalence (prevalence) varies from about 4% to about 8%; B) this figure shows the distribution of the start codons for a baculovirus library made from the inserted expression cassettes flanked by AAV ITRs. The prevalence varies from about 4% to about 9%, it being noted that the distribution properties of this library are very similar to those of a plasmid library; C) this figure shows the distribution of the initiation codons contained in the AAV library made from the baculovirus library of fig. 2B). Note that the distribution of the initiation codons is very similar to the baculovirus library, ranging from about 4% to 9%, with one exception, the prevalence of ATG initiation codons is much less than 0.5%; D) this figure shows the distribution of the initiation codon contained in the AAV vector genome in cells infected with the AAV library of fig. 2C). Note that the distribution of the start codons is now very different. CTG and GTG are the most common initiation codons, with a prevalence of about 50%. The initiation codon, ATG, represented very rarely in AAV libraries is now prevalent at about 8%, while the remaining initiation codons are prevalent at between about 1% and 3%. E) This figure combines the respective figures of fig. 2A) -D). Note the low point for the ATG codon of the AAV library and the peak points for CTG, GTG and ATG of the cell library.

FIG. 3 the selected sequences (ATG1, ATG2, etc.) were then cloned into baculovirus vectors for expression of the AAV5 capsid. The cloning of the baculovirus vectors was then analyzed by SDS-PAGE to assess the ratio of VP1: VP2: VP 3. CTG1 did not produce good clones, whereas CTG2 and GTG2 produced good clones showing the stoichiometric ratios as shown above (Lubelski et al, WO 2015137802). The TAG clone produced low titers, and the TGA clone did not appear to produce VP 1. Surprisingly, ATG1 and ATG2 produced good clones with similar stoichiometric ratios as CTG 2.

Figure 4 AAV titer assay. The relative titers of different AAV vectors carrying the SEAP reporter gene under the control of the CMV promoter were tested in Huh7 cells (a) and HeLa cells (B). With different multiplicity of infection (10)⁶、10⁵、10⁴AAV (gc)/cell at genomic copy number) infected cells and the expression of SEAP reporter was determined. The ATG1 construct gave the highest titer of the vector, whereas ATG2, CTG2 and GTG2 had similar properties, whereas TGA had significantly lower titers because it did not or hardly contained any VP1 protein. GTG1 AAV vector produced low titers of gc/ml, and therefore was not allowed to have an MOI of 10⁶The infection of (2).

FIG. 5A background schematic of ATG sequences for efficient AAV capsid protein expression. A) The upper box shows codons in the open reading frame for VP1, from left to right. The box with the VP1 start codon contains an "ATG". The lower box is out of frame with an open reading frame for VP 1. Upstream of the ATG codon is a selector start codon (start) and downstream thereof is a stop codon (stop). B) Shown is the sequence of the dominant sequence selected from the library (SEQ ID NO: 1). In an out-of-frame overlapping reading frame (OOF), the ATG initiation codon is found upstream of the in-frame reading frame for Cap. In sequences derived from the wild-type AAV5 sequence, OOF has a TGA downstream stop codon, which when translated from the OOF start codon will result in the 6 amino acid short peptide MHHGK (SEQ ID NO: 72); C) shown is the sequence of yet another out-of-frame overlapping reading frame from another upstream initiation codon. The above case has an out-of-frame CTG start codon (SEQ ID NO:2) with a stop codon further downstream of the sequence derived from AAV5 sequence (see i.a. SEQ ID NO: 70). This will result in translation of a larger protein sequence of about 158 amino acids that terminates in the TAG stop codon. The case where the CTG start codon, which also has a stop codon in the mutated sequence immediately downstream of the start codon, when translated from OOF (SEQ ID NO:9), will result in a short peptide of 4 amino acids MEIW (SEQ ID NO: 73); D) schematic representation of VP1, VP2, and VP3 capsid protein expression from constructs as shown in fig. 5A-C. The DNA contains an expression cassette with a promoter (P) and an open reading frame for the capsid protein (Cap (VP 123)). The arrow indicates the start of transcription. Transcription results in an mRNA from which the OOF protein can be translated first, and then the VP1, VP2, and VP3 capsid proteins. The OOF sequence overlaps with the VP1 translation start.

FIG. 6 is a schematic representation of various vector carrier structures for AAV library preparation. FIG. 6A: shown is the structure used in the examples, wherein expression cassettes for expression of AAV capsid proteins (grey boxes) are contained between AAV ITRs and within the baculovirus genome. The AAV produced therefrom contains a vector genome with ITRs flanked by expression cassettes. FIG. 6B: shown is a structure wherein the vector carrier (e.g., baculovirus) contains an expression cassette for parvovirus capsid protein and wherein a sequence Identifier (ID) is placed between the vector genomic ITR sequences. The AAV produced therefrom contains a vector genome with ITRs flanked by sequence markers. By identifying the sequence identifier, e.g. by sequencing, because the Cap sequence and the ID are linked, the corresponding Cap sequence can be determined, because the ID and Cap sequences are associated in a certain genome, e.g. by sequencing a baculovirus vector comprising both sequences, or because the baculovirus vector is constructed in such a way that the combination of the identifier sequence and the Cap expression sequence is known beforehand. The construct of FIG. 6C may also include a reporter gene within the parvoviral vector genome.

Definition of

As used herein, the term "operably linked" refers to the linkage of polynucleotide (or polypeptide) elements in a functional relationship. A nucleic acid is "operably linked" when the nucleic acid is placed in a functional relationship with another nucleic acid sequence. For example, a transcriptional regulatory sequence is operably linked to a coding sequence if it affects the transcription of the coding sequence. Operably linked means that the DNA sequences being linked are generally contiguous and, where necessary to join two protein coding regions, contiguous and in reading frame.

"expression control sequence" refers to a nucleic acid sequence that regulates the expression of a nucleotide sequence to which it is operably linked. An expression control sequence is "operably linked" to a nucleotide sequence when the expression control sequence controls and regulates the transcription and/or translation of the nucleotide sequence. Thus, expression control sequences may include promoters, enhancers, Internal Ribosome Entry Sites (IRES), transcription terminators, initiation codons in front of protein-encoding genes, splicing signals for introns, and stop codons. The term "expression control sequence" is intended to include (at least) sequences whose presence is designed to affect expression, and may also include additional advantageous components. For example, leader sequences and fusion partner sequences are expression control sequences. The term may also include the design of nucleic acid sequences such that in-frame and out-of-frame undesired potential start codons are removed from the sequence. It may also include the design of nucleic acid sequences such that undesired potential splice sites are removed. It includes a sequence leading to the addition of a poly A tail or polyadenylation sequence (pA), a string of adenine residues at the 3' end of the mRNA, called the polyA sequence. It can also be designed to enhance the stability of mRNA. Expression control sequences which influence the stability of transcription and translation (e.g.promoters), and sequences which influence translation (e.g.Kozak sequences), are known in insect cells. The expression control sequence may have the property to regulate the nucleotide sequence to which it is operably linked, such that a lower expression level or a higher expression level is achieved.

As used herein, the term "promoter" or "transcriptional regulatory sequence" refers to a nucleic acid fragment that functions to control the transcription of one or more coding sequences, and is located upstream in the direction of transcription relative to the transcription start site of the coding sequence, and is structurally identified by the presence of: a binding site for a DNA-dependent RNA polymerase, a transcription initiation site, and any other DNA sequence including, but not limited to, transcription factor binding sites, repressor and activator protein binding sites, and any other sequence of nucleotides known to those skilled in the art that directly or indirectly regulates the amount of transcription from a promoter. A "constitutive" promoter is a promoter that is active in most tissues under most physiological and developmental conditions. An "inducible" promoter is a promoter that is regulated physiologically or developmentally, for example by the application of a chemical inducer. A "tissue-specific" promoter is active only in a specific type of tissue or cell.

The terms "substantially identical", "substantial identity" or "substantially similar" or "substantial similarity" refer to two peptide sequences or two nucleotide sequences that, when optimally aligned, share at least a certain percentage of sequence identity as defined elsewhere herein, e.g., by the programs GAP or BESTFIT using default parameters. GAP uses Needleman and Wunsch global alignment algorithms to align two sequences over their entire length, maximizing the number of matches and minimizing the number of GAPs. Typically, GAP default parameters are used, with GAP creation penalty of 50 (nucleotides)/8 (protein) and GAP extension penalty of 3 (nucleotides)/2 (protein). For nucleotides, the default scoring matrix used was nwsgapdna, and for proteins, the default scoring matrix was Blosum62(Henikoff & Henikoff, 1992, PNAS 89, 915-919). It is clear that thymine (T) in a DNA sequence is considered equivalent to uracil (U) in an RNA sequence when the RNA sequence is said to be substantially similar or have a degree of sequence identity to the DNA sequence. Sequence alignments and percent sequence identity scores can be determined using computer programs (e.g., GCG Wisconsin Package, version 10.3, available from Accelrys Inc.,9685 Scanton Road, San Diego, CA92121-3752, USA or Windows open source software Emboss (current version 2.7.1-07)). Alternatively, the percent similarity or identity can be determined by searching a database, such as FASTA, BLAST, or the like.

The nucleotide sequence of the invention encoding a parvoviral Rep or Cap protein can also be defined by: their ability to hybridize to their respective nucleotide sequences under moderate hybridization conditions, or preferably under stringent hybridization conditions. Stringent hybridization conditions are defined herein as hybridization under conditions that allow for at least about 25 nucleotides, preferably about 50 nucleotides, 75 nucleotides, or 100 nucleotides, and most preferably about 200 or more nucleotides of a nucleic acid sequence to be washed in a solution containing about 1M salt (preferably 6x SSC) or any other solution having comparable ionic strength at a temperature of about 65 ℃, and in a solution containing about 0.1M salt or less (preferably 0.2x SSC) or any other solution having comparable ionic strength at 65 ℃. Preferably, the hybridization is performed overnight, i.e., at least 10 hours, and preferably at least 2 changes of washing solution for at least 1 hour. These conditions will generally allow for specific hybridization of sequences having about 90% or greater sequence identity.

Mild conditions are defined herein as hybridization under conditions that allow for at least about 50 nucleotides, preferably about 200 nucleotides or more, of a nucleic acid sequence to be washed in a solution comprising about 1M salt (preferably 6xSSC) or any other solution with comparable ionic strength at a temperature of about 45 ℃ and in a solution comprising about 1M salt (preferably 6xSSC) or any other solution with comparable ionic strength at room temperature. Preferably, the hybridization is performed overnight, i.e., at least 10 hours, and preferably at least 2 changes of washing solution for at least 1 hour. These conditions will generally allow for specific hybridization of sequences having up to 50% sequence identity. One skilled in the art will be able to modify these hybridization conditions in order to specifically identify sequences having an identity varying between 50% and 90%.

Detailed Description

The present invention relates to the use of an animal parvovirus, in particular a dependent virus (dependently), such as infectious human or simian AAV and components thereof (e.g. an animal parvovirus genome) as a vector for introducing and/or expressing a nucleic acid in a mammalian cell. In particular, the invention relates to the improvement of the infectivity of the parvoviral vector when produced in insect cells.

Viruses of the parvoviridae family are small DNA animal viruses. The parvoviridae can be divided into two subfamilies: the Parvovirinae (Parvovirinae) that infects vertebrates and the Densovirinae (Densvirinae) that infects insects. Members of the parvovirinae are referred to herein as parvoviruses and include dependovirus. It can be concluded from the name of its genus that members of the dependent viruses are unique in that they usually require co-infection with a helper virus such as adenovirus or herpes virus for productive infection in cell culture. The genus dependovirus includes AAV, which normally infects humans (e.g., serotypes 1, 2, 3A, 3B, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13) or primates (e.g., serotypes 1 and 4), as well as related viruses that infect other warm-blooded animals (e.g., bovine, canine, equine, and ovine adeno-associated viruses). More information on parvoviruses and other members of The Parvoviridae family is described in Kennethi I.Berns, "Parvoviridae: The Viruses and The ir Replication," Chapter69in Fields Virology (third edition, 1996). For convenience, the invention is further explained and described herein by reference to AAV. However, it is to be understood that the invention is not limited to AAV, but is equally applicable to other parvoviruses.

The genomic structure of all known AAV serotypes is very similar. The genome of AAV is a linear, single-stranded DNA molecule that is less than about 5000 nucleotides (nt) in length. Inverted Terminal Repeats (ITRs) flank the unique coding nucleotide sequences of the nonstructural replication (Rep) and structural (VP) proteins. The VP proteins (VP1, -2, and-3) form the capsid. The ends 145nt are self-complementary and are organized such that an energy-stable intramolecular duplex can be formed, which forms a T-shaped hairpin. These hairpin structures function as origins of viral DNA replication and serve as primers for the cellular DNA polymerase complex. Upon wtAAV infection of mammalian cells, the Rep genes (i.e., Rep78 and Rep52) are expressed from the P5 promoter and P19 promoter, respectively, and both Rep proteins are functional in replication of the viral genome. Splicing events in the Rep ORF result in the expression of the actual four Rep proteins (i.e., Rep78, Rep68, Rep52, and Rep 40). However, it has been shown that in mammalian cells, unspliced mRNA encoding the Rep78 and Rep52 proteins is sufficient for production of AAV vectors. Also in insect cells, the Rep78 and Rep52 proteins are sufficient for production of AAV vectors. The three capsid proteins VP1, VP2, and VP3 are expressed from the p40 promoter from a single VP reading frame. For capsid protein production, the infection of wtAAV in mammalian cells relies on a combination of selective use of two splice acceptor sites and suboptimal utilization of the ACG start codon of VP 2.

In insect cells, expression of transcripts (i.e., mRNA) having AAV open reading frames encoding VP1 (with an AUG start codon), VP2, and VP3 proteins, does not typically produce VP1, VP2, and VP3 capsid proteins in a ratio of about 1:1:10 and in amounts that result in a potent AAV. A titer, as defined herein, is the ability of an AAV vector to transfer its vector genome to a target cell and allow for efficient expression of the transgene. The present inventors have now surprisingly found that AAV capsids can be produced in insect cells with high efficiency from expression constructs encoding transcripts for VP1, VP2, and VP3 proteins, wherein VP1 is translated from the AUG start codon.

The expression constructs identified in the present invention allow for the efficient production of large numbers of high titer AAV gene therapy vectors for medical treatment in insect cells. Such vectors are at least similar if there is no improvement in the potency of the AAV gene therapy vector as generated from the selectable initiation codon (e.g. CTG or GTG) (fig. 4). It is understood that, with respect to nucleic acid sequences, it may be listed as DNA sequences (listed as A, T, C and G) or RNA sequences (listed as A, U, C and G). It is understood that an expression construct may generally refer to a DNA sequence, while an expressed nucleotide sequence refers to an RNA sequence, i.e. an mRNA transcribed or expressed from the expression construct.

The constructs of the invention encode an additional out of frame start codon 5' to the VP1 start codon, which apparently results in a reduction in translation initiation at the VP1 start codon, which allows for further translation of both VP2 and VP3 in sufficient amounts. Without being limited by theory, such an out-of-frame 5' start codon results in interference with transcription initiation at the VP1 AUG start codon and allows for a pseudoleaky ribosome scan similar to that which occurs in wild-type AAV. Without being limited by theory, synthesis of short peptides from these alternative initiation codons (e.g., termination of translation of the out-of-frame reading frame preceding the VP2 coding sequence) may allow the ribosome to continue to scan downstream of the VP1 AUG initiation codon or to reinitiate it, which allows for translation of VP2 and VP3 from the same transcript.

Such constructs provide at least similar, if not improved, production of AAV capsids with good titer in insect cells compared to AAV capsids produced by the prior art. Advantageously, when used to generate expression constructs for insect cells, such constructs may allow for unmodified VP1, VP2, and VP3 nucleotide sequences as found in wild-type viruses. The constructs according to the invention may allow the amino acid sequence of the VP1, VP2 and VP3 capsid proteins to be substantially identical or identical to the capsid proteins found in the wild type virus. Thus, this expression strategy is generally applicable to any parvoviral or AAV vector construct, and may not require further tailoring of the 5' sequence or the sequence of the AAV capsid open reading frame.

Accordingly, in a first aspect of the invention there is provided a nucleic acid construct comprising an expression control sequence for expressing a nucleotide sequence comprising an open reading frame in an insect cell, wherein the open reading frame sequence encodes:

i) adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP 3; and

ii) ATG translation initiation codon for VP 1;

the nucleotide sequence includes a selector start codon upstream of the open reading frame, the selector start codon being out-of-frame with respect to the open reading frame.

It is understood that expression of a nucleotide sequence according to the invention is related to the expressed mRNA. Thus, the selector initiation codon will be included in the mRNA, i.e. it is included in the sequence 5 'of the open reading frame encoding the capsid protein, and it is 3' of the transcription initiation site of the nucleic acid construct. Thus, the alternative reading frame is 5' of the VP1 AUG codon included in the expressed mRNA. It is understood that an open reading frame according to the present invention is understood as a single open reading frame, i.e. the sequences encoding the capsid proteins VP1, VP2 and VP3 are overlapping. In other words, the VP2 and VP3 proteins are encoded by sequences identical to the VP1 sequence. Such open reading frames may be continuous open reading frames, but may also be discontinuous, e.g. containing intron sequences. Preferably, the open reading frames from which VP1, VP2, and VP3 are translated are a contiguous single open reading frame, wherein no other transcript from which capsid proteins can be translated in the insect is transcribed (e.g., when one transcript encodes VP1 and the other transcript encodes VP2, and still the other transcript encodes VP 3).

Preferably, the out-of-frame start codon is selected from the group consisting of: CUG, ACG, AUG, UUG, CUC and CUU. More preferably, the selector start codon is selected from AUG or CUG. Most preferably, the selector initiation codon is AUG. As shown in the examples section, the sequence with the most prevalent AUG initiation codon contains mainly an out-of-frame initiation codon. The upstream out-of-frame start codon is mainly a relatively strong codon, such as UUG, CUG, GUG, AUG and ACG. Weak initiation codons, such as CUC and CUU, were also observed. Most commonly and most preferably AUG is used as the out of frame selective initiation codon.

The selectable initiation codon can be the start of a selectable open reading frame. Thus, a selectable initiation codon is understood to include a codon from which a ribosome can initiate translation. Sometimes such sequences may not be allowed to function as a start codon when the start codon is, for example, near the 5' cap end of the mRNA. It is understood that due to the genetic code, where a triplet (triplet) encodes an amino acid, a nucleic acid sequence can be translated into three different amino acid sequences depending on where the translation is initiated and terminated. The out-of-frame selective initiation codon is upstream of the VP1 AUG initiation codon, and preferably the genetic code following the selective initiation codon is such that termination of translation occurs such that the ribosome does not initiate or is prevented from initiating translation from the VP1 AUG initiation codon. Also, without being limited by theory, an out-of-frame selective initiation codon upstream of the VP1 AUG initiation codon allows for initiation of translation from mRNA. Preferably, the selectable open reading frame terminates downstream of the VP1 AUG start codon. For example, when the VP1 AUG start codon will be followed by a, the UGA triplet in the AUGA sequence encodes a stop codon. Thus, preferably, the selectable open reading frame starting upstream from the selector start codon encompasses the VP1 AUG start codon.

Thus, in yet another embodiment according to the present invention, there is provided a nucleic acid construct for expressing a nucleotide sequence comprising an open reading frame in an insect cell, wherein the open reading frame sequence encodes adeno-associated virus (AAV) capsid proteins VP1, VP2 and VP3 and an AUG translation initiation codon for VP1, wherein the nucleotide sequence comprises a selective open reading frame starting with a selective initiation codon, wherein the selective open reading frame encompasses the AUG translation initiation codon for VP 1.

The alternative open reading frame preferably starts at up to 100, 90, 80, 70, 60, 50, 40, 30, 20 or 10 nucleotides 5' of the VP1 AUG start codon and terminates thereafter. The alternative open reading frame starts 5' from the VP1 AUG start codon and terminates at most 500, 400, 300, 200, 150, 100, 90, 80, 70, 60, 50, 40, 30, 20, or 10 nucleotides thereafter. The alternative open reading frame may start at most 50 nucleotides 5' of the VP1 AUG start codon and terminate at most 500 nucleotides thereafter. The alternative open reading frame may start at up to 40 nucleotides 5' of the VP1 AUG start codon and terminate at up to 200 nucleotides thereafter. The alternative open reading frame may also start at up to 30 nucleotides 5' of the VP1 AUG start codon and terminate at up to 50 nucleotides thereafter. The alternative open reading frame may start at up to 10 nucleotides of the VP1 AUG start codon and terminate at up to 20 nucleotides thereafter. In an alternative embodiment, the alternative open reading frame terminates before the start codon of VP3, preferably before the start codon of VP 2. For example, such alternative open reading frames, as shown in the examples, start 4 nucleotides upstream and terminate 14 nucleotides thereafter, or start 8 nucleotides upstream and terminate 4 or more nucleotides thereafter.

Preferably, such alternative open reading frames may be included in the DNA sequences encoding the adeno-associated virus (AAV) capsid proteins VP1, VP2 and VP3, including the sequence upstream of the VP1ATG start codon sequence encoded by nucleotides 105 to 155 of the DNA sequence of SEQ ID NO: 70. The sequence upstream of the ATG initiation codon is transcribed into RNA. Such alternative open reading frames may also be included in the DNA sequences encoding the adeno-associated virus (AAV) capsid proteins VP1, VP2 and VP3, including the sequence upstream of the VP1ATG initiation codon sequence encoded by nucleotides 1-155 of the DNA sequence of SEQ ID NO: 70. The upstream sequence encodes the polyhedrin promoter and 5' leader sequence upstream of the ATG VP1 start codon (105-155).

Thus, preferably, the alternative open reading frame of the present invention as described above is translated into a peptide in insect cells. In one embodiment, the peptide has a length of at least 4 amino acids, at least 5 amino acids, at least 6 amino acids. In one embodiment, the translated amino acid sequence comprises SEQ ID NO:72 or SEQ ID NO:73 or by SEQ ID NO:72 or SEQ ID NO: 73. In another embodiment, the peptide has a length of up to 200, 150, 100, 50, 40, 30, 20, or 10 amino acids. In yet another embodiment, the nucleic acid construct encoding said alternative open reading frame according to the invention is translated into a peptide ranging from 2 to 200 amino acids, 2 to 100 amino acids, 2 to 50 amino acids or preferably 2 to 10 amino acids in length. Thus, a nucleic acid construct according to the invention as described herein, which comprises the alternative open reading frame following the alternative initiation codon, encodes a peptide. The length of the peptide may depend on the sequence following the start codon of VP1, i.e. the sequence encoding VP1, which may for example be derived from AAV sequences derived from nature, or from synthetic or artificial AAV capsid sequences (e.g. codon optimized or mutated variants with improved properties). Thus, the length depends on where the stop codon (TGA, TAA, TAG) occurs in the out-of-frame reading frame starting from the selector initiation codon upstream of the VP1ATG initiation codon. The sequence downstream of the start codon can be mutated to introduce a stop codon that is in-frame with the out-of-frame upstream start codon. In this way, the length of the peptide can be purposefully selected. Thus, an out-of-frame stop codon may be introduced which does not introduce a change in amino acid sequence for the VP1 coding sequence, in other words a silent mutation in the VP1 reading frame. The introduced out-of-frame stop codon may be introduced by one, two or three point mutations in three consecutive nucleic acids in the reading frame. A triplet sequence (i.e. TGA, TAA or TAG) may also be inserted within the VP1 coding sequence, which may result in the insertion of one amino acid for the length of the coding sequence and may result in additional amino acid changes to the VP1 coding sequence (i.e. one triplet of the VP1 coding sequence is changed to two triplets by the insertion of an out-of-frame stop codon).

In another embodiment, there is provided a nucleic acid construct comprising an expression control sequence for expressing a nucleotide sequence comprising an open reading frame in an insect cell, wherein the open reading frame sequence encodes:

i) adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP 3; and

ii) the AUG translation initiation codon for VP 1;

wherein the nucleotide sequence directly includes a nucleotide sequence at nucleotides 1-8 upstream of VP1 AUG, the nucleotide sequence being selected from the group consisting of SEQ ID NO. 32-62. It is understood that SEQ ID NO.32-62 refers to an RNA sequence and thus the nucleic acid construct will have the corresponding DNA sequence encoding said RNA sequence, for example as set out in SEQ ID NO 1-31. Preferably, the nucleotide sequence includes a G nucleotide immediately downstream of VP1 AUG. More preferably, the nucleic acid construct according to the invention comprises a sequence selected from the group consisting of SEQ ID NO.1-31 encoding the start codon of VP1, wherein said VP1 start codon corresponds to positions 9-11 of said SEQ ID NO 1-31. Most preferred are sequences derived from SEQ ID NO.1 and SEQ ID NO.32, i.e.preferably nucleotides 1 to 8 thereof, having a G preferably directly adjacent to VP1ATG, more preferably the entire sequence encoding SEQ ID NO. 1.

In yet another embodiment, there is provided a nucleic acid construct according to the invention, wherein the second codon of the open reading frame of VP1 encodes an amino acid residue selected from the group consisting of: alanine, glycine, valine, aspartic acid and glutamic acid. This second amino acid residue may be derived from a codon inserted between the start codon and a second codon derived from, for example, the wild-type AAV VP1 sequence, or the second codon of the VP1 nucleotide sequence may be a mutated codon (e.g., by mutating the nucleic acid immediately following the VP1ATG codon to G). Most preferably, the second codon of VP1 encodes valine. More preferably, the second codon is selected from the group consisting of GUA, GUC, GUU, GUG, preferably the second codon is GUA. The open reading frame optionally includes one or more codons further encoding additional amino acid residues after the second codon, e.g., codons for 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 additional amino acids, but preferably less than 60, 50, 40, 35, 30, 25, 20, 19, 18, 17, 16, 15, or 14 additional amino acid residues. As will be readily appreciated, codons encoding additional amino acid residues are in-frame with the open reading frame of the capsid protein.

Thus, in one embodiment, AAV vectors are provided that include a VP1 capsid protein having a valine at position 2 of VP1, by modifying, for example, position 2 of the wild-type VP1 capsid protein sequence or by inserting a valine codon between

positions

1 and 2 of the wild-type VP1 capsid protein sequence, or because a VP1 capsid protein found or selected in nature already includes a valine at position 2. Such capsids, preferably produced in insect cells, may be particularly useful in medical treatment as described herein.

In one embodiment, if the open reading frame is compared to a wild-type capsid protein, the open reading frame encoding the capsid protein further comprises a codon encoding one or more amino acid residues inserted between the ATG translation start codon of VP1 and the codon encoding the amino acid residue immediately 3' of the start codon in the corresponding wild-type capsid protein. For example, the open reading frame comprises codons for 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 additional amino acid residues compared to the corresponding wild-type capsid protein. Preferably, the open reading frame comprises codons for less than 60, 50, 40, 35, 30, 25, 20, 19, 18, 17, 16, 15 or 14 additional amino acid residues compared to the corresponding wild type capsid protein. As will be readily appreciated, codons encoding additional amino acid residues are in-frame with the open reading frame of the capsid protein. Of these codons encoding additional amino acid residues, the first codon, i.e. the codon immediately 3' of the suboptimal translation initiation codon, encodes an amino acid residue selected from the group consisting of: alanine, glycine, valine, aspartic acid and glutamic acid. Thus, if there is only one additional codon between the translation initiation codon and the codon encoding the amino acid residue corresponding to residue 2 of the wild-type sequence, the additional codon encodes an amino acid residue selected from the group consisting of: alanine, glycine, valine, aspartic acid and glutamic acid. If there is more than one additional codon between the translation initiation codon and the codon encoding amino acid residue 2 of the wild-type sequence, then the codon immediately following the translation initiation codon encodes an amino acid residue selected from the group consisting of: alanine, glycine, valine, aspartic acid and glutamic acid. Preferably, the additional amino acid residue immediately following (i.e., at the 3' end of) the suboptimal translation initiation codon is valine. In other words, in a preferred embodiment of the invention, the codon immediately following the suboptimal translation initiation codon encodes a valine.

The sequence encoding the AAV capsid protein in step a) may be a capsid sequence found in nature, such as AAV1-AAV13, the nucleotide and amino acid sequence of which is described in Lubelski et al, WO2015137802 as SEQ ID NO: 13-38, which are hereby incorporated by reference in their entirety. Thus, the nucleic acid construct according to the invention may comprise the complete open reading frame of the AAV capsid protein as disclosed by Lubelski et al, WO 2015137802. Alternatively, the sequence may be artificial, e.g., the sequence may be in heterozygous form, or may be codon optimized, e.g., by codon usage of AcmNPv or spodoptera rugiperda. For example, the capsid sequence may consist of the VP2 and VP3 sequences of AAV1, while the remainder of the VP1 sequences are of AAV 5. A preferred capsid protein is AAV5, preferably as provided in SEQ ID NO:22 or AAV8, preferably as provided in SEQ ID NO:28 as set forth in Lubelski et al WO 2015137802. Thus, in a preferred embodiment, the AAV capsid protein is a capsid protein of AAV serotype 5 or AAV serotype 8 that has been modified according to the invention. More preferably, the AAV capsid protein is a capsid protein of AAV serotype 5 that has been modified according to the invention. It is understood that the exact molecular weight of the capsid protein and the exact position of the translation initiation codon may vary between different parvoviruses. However, one skilled in the art would know how to identify the corresponding positions in the nucleotide sequence from other parvoviruses than AAV-5. Alternatively, the sequences encoding the AAV capsid proteins are artificial sequences, e.g., as a result of directed evolution experiments. This may include generating capsid libraries by DNA shuffling, error-prone PCR, bioinformatics rational design, site-saturation mutagenesis. The resulting capsids are based on existing serotypes, but contain a variety of amino acid or nucleotide changes that improve the characteristics of such capsids. The resulting capsids may be a combination of portions of existing serotypes, "shuffled capsids," or contain entirely new changes, i.e., additions, deletions, or substitutions of one or more amino acids or nucleotides, arranged in a group or distributed throughout the length of a gene or protein. See, e.g., schafer and Maheshri; proceedings of the 26th annual International Conference of the IEEE EMBS San Francisco, CA, USA; september 1-5,2004, pages 3520-3523; asuri et al (2012) Molecular Therapy 20(2): 329-3389; lisowski et al (2014) Nature 506(7488):382-386, incorporated herein by reference.

In a preferred embodiment of the invention, the open reading frame encoding the VP3 capsid protein is initiated with a non-canonical (non-canonical) translation initiation codon selected from the group consisting of: ACG, ATT, ATA, AGA, AGG, AAA, CTG, CTT, CTC, CTA, CGA, CGC, TTG, TAG and GTG. Preferably, the non-canonical translation initiation codon is selected from the group consisting of: GTG, CTG, ACG, TTG, more preferably, the non-canonical translation initiation codon is CTG.

The nucleotide sequence of the present invention for expressing an AAV capsid protein preferably further comprises at least one modification of the nucleotide sequence encoding an AAV VP1 capsid protein selected from the group consisting of G at nucleotide 12, a at nucleotide 21 and C at nucleotide 24 of the VP1 open reading frame, wherein the nucleotide position corresponds to the nucleotide position of the wild-type nucleotide sequence. A "potential/potential wrong initiation site" or a "potential/potential wrong translation initiation codon" is understood herein to mean an in-frame ATG codon located in the coding sequence of the capsid protein. The elimination of a possible incorrect initiation site for translation within the VP1 coding sequence of other serotypes, as well as the elimination of putative splice sites that may be recognized in insect cells, will be well understood by those skilled in the art. For example, recombinant AAV5 does not require a modification at nucleotide 12, because nucleotide T does not result in a spurious ATG codon. Various modifications of wild-type AAV sequences for proper expression in insect cells are achieved by applying well-known genetic engineering techniques as described in Sambrook and Russell (2001) in molecular cloning, A Laboratory Manual (3rd edition), Cold Spring Harbor Laboratory Press, New York. A number of further modifications of the VP-encoding region are known to the skilled person, which modifications may increase the yield of VP and virion or have other desired effects, such as altering the tropism of the virion or reducing the antigenicity of the virion. Such modifications are within the scope of the present invention.

Preferably, the nucleotide sequence of the invention encoding an AAV capsid protein is operably linked to an expression control sequence for expression in insect cells. Thus, in a second aspect, the present invention relates to a nucleic acid construct comprising a nucleic acid molecule according to the present invention, wherein a nucleotide sequence encoding an open reading frame for an adeno-associated virus (AAV) capsid protein is operably linked to an expression control sequence for expression in an insect cell. These expression control sequences will include at least a promoter active in insect cells. Techniques known to those skilled in the art for expressing foreign genes in insect host cells can be used to practice the present invention. Methods of molecular engineering and Methods of polypeptide expression in Insect cells are described, for example, in Summers and Smith.1986.A Manual of Methods for bacterial Vectors and Instrument Current Vectors, Texas Agricultural Experimental Station Bull.No.7555, CollegeStation, Tex.; luckow.1991.in Prokop et al, Cloning and Expression of heterologous Genes in institute Cells with Baculoviral Vectors' recombining DNAtechnology and Applications, 97-152; king, l.a.and r.d.pos, 1992, thebaculoviral expression system, Chapman and Hall, United Kingdom; o' Reilly, D.R., L.K.Miller, V.A.Luckow,1992, Baculoviral Expression Vectors A laboratory Manual, New York; freeman and Richardson, c.d.,1995, bactovovirus expression protocols, Methods in Molecular Biology, volume 39; US4,745,051; US 2003148506; and WO 03/074714. A particularly suitable promoter for transcription of a nucleotide sequence of the invention encoding an AAV capsid protein is, for example, the polyhedral promoter (polH), such polH promoter being provided in SEQ ID NO:70 (or as set forth in SEQ ID NO:53, and a shortened version thereof SEQ ID NO: 54, in WO2015137802 by Lubelski et al). However, other promoters active in insect cells are known in the prior art and may be selected according to the invention, such as the polyhedrin (polH) promoter, the p10 promoter, the p35 promoter, the 4xHsp27 EcRE + minor Hsp70 promoter, the deltaE1 promoter, the E1 promoter or the IE-1 promoter, as well as the other promoters described in the above-mentioned references.

Preferably, the nucleic acid construct for expressing the AAV capsid protein in an insect cell is an insect cell-compatible vector. An "insect cell-compatible vector" or "vector" is understood to be a nucleic acid molecule capable of generative transformation or transfection of an insect or insect cell. Exemplary biological vectors include plasmids, linear nucleic acid molecules, and recombinant viruses. Any vector may be used as long as it is insect cell-compatible. The vector may be integrated into the genome of the insect cell, but the presence of the vector in the insect cell need not be permanent and also includes transient episomal vectors. The vector may be introduced by any known means, for example by chemical treatment of the cells, electroporation or infection. In a preferred embodiment, the vector is a baculovirus, a viral vector or a plasmid. In a more preferred embodiment, the vector is a baculovirus, i.e. the construct is a baculovirus vector. Baculovirus vectors and methods of use thereof are described in the molecular engineering references cited above for insect cells.

In a third aspect, the present invention relates to an insect cell comprising the nucleic acid construct of the invention described above. Any insect cell that allows replication of AAV and that can be maintained in culture can be used according to the invention. For example, the cell line used may be from Spodoptera frugiperda (Spodoptera frugiperda), a Drosophila cell line or a mosquito cell line, for example derived from Aedes albopictus (Aedes albopictus). Preferred insect cells or cell lines are cells from insect species susceptible to baculovirus infection, including, for example, expressSF + from Invitrogen

Drosophila Schneider 2(S2) cells, Se301, SeIZD2109, SeUCR1, Sf9, Sf900+, Sf21, BTI-TN-5B1-4, MG-1, Tn368, HzAm1, Ha2302, Hz2E5 and High Five.

Preferred insect cells according to the invention further comprise: (a) a second nucleotide sequence comprising at least one AAV Inverted Terminal Repeat (ITR) nucleotide sequence; (b) a third nucleotide sequence comprising a Rep52 or Rep40 coding sequence operably linked to expression control sequences for expression in an insect cell; and (c) a fourth nucleotide sequence comprising a Rep78 or Rep68 coding sequence operably linked to expression control sequences for expression in an insect cell.

In the context of the present invention, "at least one AAV ITR nucleotide sequence" is understood to refer to a palindromic sequence, which includes most of the complementary, symmetrically arranged sequences, also referred to as the "a", "B" and "C" regions. The ITRs function as origins of replication, sites that have a "cis" role in replication, i.e., recognition sites for trans-acting replication proteins (e.g., Rep78 or Rep68) that recognize palindromes and specific sequences within the palindromes. One exception to the symmetry of the ITR sequence is the "D" region of the ITR. It is unique (no complement within one ITR). Nicking of single-stranded DNA occurs at the junction of the A-and D-regions. Which is the region where new DNA synthesis starts. The D region is usually located on one side of the palindrome and provides directionality to the nucleic acid replication steps. AAV, which replicates in mammalian cells, typically has two ITR sequences. However, it is possible to engineer the ITRs so that the binding sites are on both strands of the A region and the D region is symmetrically one on each side of the palindrome. On a double-stranded circular DNA template (e.g., a plasmid), Rep 78-assisted or Rep 68-assisted nucleic acid replication then proceeds in both directions, and a single ITR is sufficient for AAV replication of the circular vector. Thus, an ITR nucleotide sequence may be used in the context of the present invention. However, it is preferred to use two or another even number of regular ITRs. Most preferably, two ITR sequences are used. In view of the safety of viral vectors, it would be desirable to be able to construct viral vectors that are incapable of further propagation after initial introduction into cells. Such a safety mechanism for limiting undesired vector proliferation in a recipient may be provided by using rAAV with a chimeric ITR as described in US 2003148506. In a preferred embodiment, the nucleotide sequences encoding the parvovirus VP1, VP2 and VP3 capsid proteins comprise at least one in-frame insertion of a sequence encoding an immune escape repeat, as described in WO 2009/154452. This leads to the formation of virions called self-complementary or monomeric duplex parvoviruses. In a preferred embodiment, the sequences encoding the parvovirus VP1, VP2, and VP3 capsid proteins comprise a monomeric duplex or a self-complementary genome. To prepare a monomeric duplex AAV vector, AAV Rep proteins and AAV capsid proteins are expressed in an insect cell according to the invention and in the presence of a vector genome comprising at least one AAV ITR, wherein Rep52 and/or Rep40 protein expression is increased relative to Rep78 and/or Rep68 protein expression. Monomeric, double-stranded AAV vectors can also be prepared by expressing AAV Rep proteins and AAVCap proteins in insect cells in the presence of a vector genomic construct flanked by at least one AAV ITR, wherein the nicking activity of Rep78 and/or Rep60 is reduced relative to the helicase/encapsidation activity of Rep52 and/or Rep40, as described, for example, in WO 2011/122950.

The number of vectors or nucleic acid constructs used is not limited in the present invention. For example, one, two, three, four, five, six, or more vectors can be employed to produce AAV in an insect cell according to the invention. If six vectors are employed, one vector encodes AAV VP1, another encodes AAV VP2, yet another vector encodes AAV VP3, yet another vector encodes Rep52 or Rep40, and Rep78 or Rep68 is encoded by the other vector, and the final vector includes at least one AAV ITR. Additional vectors may be employed to express, for example, Rep52 and Rep40, as well as Rep78 and Rep 68. If less than six vectors are used, the vectors may include at least one AAV ITR and various combinations of VP1, VP2, VP3, Rep52/Rep40, and Rep78/Rep68 coding sequences. Preferably, two vectors or three vectors are used, with two vectors as described above being more preferred. If two vectors are used, preferably the insect cell comprises: (a) a first nucleic acid construct for expressing an AAV capsid protein as defined above, the construct further comprising third and fourth nucleotide sequences as defined in (b) and (c) above, said third nucleotide sequence comprising a Rep52 or Rep40 coding sequence operably linked to at least one expression control sequence for expression in an insect cell, and said fourth nucleotide sequence comprising a Rep78 or Rep68 coding sequence operably linked to at least one expression control sequence for expression in an insect cell; and (b) a second nucleic acid construct comprising a second nucleotide sequence as defined in (a) above, which comprises at least one AAV ITR nucleotide sequence. If three vectors are used, preferably the same configuration as used for two vectors is used, except that separate vectors are used for expressing capsid proteins and for expressing Rep52, Rep40 Rep78 and Rep68 proteins. The sequences on each vector may be in any order relative to each other. For example, if a vector includes ITRs and an ORF that includes nucleotide sequences encoding a VP capsid protein, the VPORF can be located on the vector such that upon replication of the DNA between the ITR sequences, the VP ORF is replicated or not replicated. As another example, the Rep coding sequence and/or the ORF including the nucleotide sequence encoding the VP capsid protein can be on the vector in any order. It is understood that the second, third and further nucleic acid constructs are preferably insect cell-compatible vectors, preferably baculovirus vectors as described above. Alternatively, in the insect cell of the present invention, one or more of the first, second, third and fourth nucleotide sequences and optionally further nucleotide sequences may be stably integrated into the genome of the insect cell. One of ordinary skill in the art knows how to stably introduce nucleotide sequences into the genome of an insect and how to identify cells having such nucleotide sequences in the genome. Incorporation into the genome can be facilitated, for example, by using a vector that includes a nucleotide sequence that is highly homologous to a region of the insect genome. The use of specific sequences, such as transposons, is another way to introduce nucleotide sequences into the genome.

Thus, in a preferred embodiment, the insect cell according to the invention comprises: (a) a first nucleic acid construct according to the invention, wherein said first nucleic acid construct further comprises a third and a fourth nucleotide sequence as defined above; and (b) a second nucleic acid construct comprising a second nucleotide sequence as defined above, wherein said second nucleic acid construct is preferably an insect cell-compatible vector, more preferably a baculovirus vector.

In a preferred embodiment of the invention, the second nucleotide sequence present in the insect cell of the invention, i.e. the sequence comprising at least one AAV ITR, further comprises at least one nucleotide sequence encoding a gene product of interest, preferably for expression in a mammalian cell, wherein preferably the at least one nucleotide sequence encoding a gene product of interest is incorporated into the genome of an AAV produced in the insect cell. Preferably, at least one nucleotide sequence encoding a gene product of interest is a sequence for expression in a mammalian cell. Preferably, the second nucleotide sequence comprises two AAV ITR nucleotide sequences and wherein the at least one nucleotide sequence encoding a gene product of interest is located between the two AAV ITR nucleotide sequences. Preferably, a nucleotide sequence encoding a gene product of interest (for expression in a mammalian cell) will be incorporated into the AAV genome produced in an insect cell if the nucleotide sequence encoding the gene product of interest (for expression in a mammalian cell) is located between two regular ITRs, or on one side of an ITR engineered with two D regions. Thus, in a preferred embodiment, the invention provides an insect cell according to the invention, wherein the second nucleotide sequence comprises two AAV ITR nucleotide sequences and wherein the at least one nucleotide sequence encoding a gene product of interest is located between the two AAV ITR nucleotide sequences.

Typically, the gene product of interest, including the ITRs, is 5000 nucleotides (nt) in length or less. In another embodiment, the AAV vector of the invention can be used to express oversized DNA, i.e., greater than 5000nt in length, in vitro or in vivo. Oversized DNA is herein understood to be DNA that exceeds the maximum AAV packaging limit by 5 kbp. Thus, production of AAV vectors capable of producing recombinant proteins typically encoded by genomes larger than 5.0kb is also feasible. For example, the present inventors have generated rAAV5 vectors containing a partial, one-way packaged fragment of hFVIII in insect cells. The total size of the vector genome, encompassing at least 5.6kb, was packaged into two populations of AAV5 particles containing FVIII fragments. These variant AAV5-FVIII vectors were shown to drive expression and secretion of active FVIII. This was demonstrated in vitro experiments in which an AAV vector comprising a gene product of interest encoding factor VIII resulted in the production of active FVIII protein upon infection of Huh7 cells. Likewise, caudal vein delivery of raav.fviii in mice results in the production of active FVIII protein. Molecular analysis of the encapsidation product clearly showed that the 5.6kbp FVIII expression cassette was not entirely encapsidated in AAV particles. Without wishing to be bound by any theory, we hypothesize that the + and-DNA strands of the encapsidated molecule show a lack of 5' end. This is consistent with the previously reported one-way (starting from the 3' end) packaging mechanism operating under the "head-complete principle" (see, e.g., Wuet al [2010] Molecular Therapy18(1): 80-86; Dong et al [2010] Molecular Therapy18(1): 87-92; Kapranov et al [2012] Human Gene Therapy 23: 46-55; and in particular Lai et al [2010] Molecular Therapy18(1): 75-79. although only about 5kb of the entire 5.6kb vector genome is encapsidated, the vector is of significant valency and causes expression of activity.

Thus, the second nucleotide sequence as defined herein above may comprise a nucleotide sequence encoding at least one "gene product of interest" for expression in a mammalian cell, positioned such that it will be incorporated into the AAV genome which replicates in an insect cell. Any nucleotide sequence may be incorporated for later expression in a mammalian cell transfected with an AAV produced according to the invention, so long as the construct remains within the packaging capacity of the AAV virion. The nucleotide sequence may, for example, encode a protein which may express an RNAi agent, i.e. an RNA molecule capable of RNA interference, such as shRNA (short hairpin RNA) or siRNA (short interfering RNA). "siRNA" refers to small interfering RNAs that are short-length double-stranded RNAs that are non-toxic in mammalian cells (Elbashir et al, 2001, Nature)411:494-98；Caplen et al.,2001,Proc.Natl.Acad.Sci.USA98:9742-47). In a preferred embodiment, the second nucleotide sequence may comprise two nucleotide sequences and each encodes a gene product of interest for expression in a mammalian cell. Each of the two nucleotide sequences encoding the product of interest is located such that it is incorporated into a rAAV genome that replicates in an insect cell.

The product of interest for expression in mammalian cells can be a therapeutic gene product. The therapeutic gene product may be a polypeptide, or an RNA molecule (siRNA) or other gene product, when expressed in a target cellFor a desired therapeutic effect, e.g. elimination of an undesired activity, e.g. elimination of infected cells, or complementation of a genetic defect, e.g. causing a defect in the enzymatic activity. Examples of therapeutic polypeptide gene products include CFTR, factor IX, lipoprotein lipase (LPL, preferably LPLS 447X; see WO 01/00220), apolipoprotein A1, uridine diphosphate glucuronosyltransferase (UGT), retinitis pigmentosa GTPase regulatory interaction protein (RP-GRIP), cytokines or interleukins such as IL-10, dystrophin, PBGD, NaGLU, Treg167, Treg289, EPO, IGF, IFN, GDNF, XP FO 3, factor VIII, VEGF, AGXT and insulin. Alternatively or additionally as a second gene product, the second nucleotide sequence as defined herein above may comprise a nucleotide sequence encoding a polypeptide as a marker protein for assessing cell transformation and expression. Suitable marker proteins for this purpose are for example the fluorescent protein GFP, and the selectable marker genes HSV thymidine kinase (for selection on HAT medium), bacterial hygromycin B phosphotransferase (for selection on hygromycin B), Tn5 aminoglycoside phosphotransferase (for selection on G418) and dihydrofolate reductase (DHFR) (for selection on methotrexate), CD20, the low affinity nerve growth factor gene. In the document Sambrook and Russel (2001) "Molecular Cloning: A Laboratory Manual (3)^rdOrigin for obtaining these marker genes and methods of using them are provided in edition), Cold Spring Harbor Laboratory Press, New York. Furthermore, the second nucleotide sequence as defined herein above may include a nucleotide sequence encoding a polypeptide that can serve as a fail-safe mechanism that allows cells transduced with the rAAV of the invention to cure a subject, if deemed necessary. Such nucleotide sequences, commonly referred to as suicide genes, encode proteins capable of converting a prodrug into a toxic substance capable of killing the transgenic cells in which the protein is expressed. Suitable examples of such suicide genes include, for example, the E.coli cytosine deaminase gene, or one of the thymidine kinase genes from Herpes Simplex Virus (Herpes Simplex Virus), Cytomegalovirus (Cytomegalovirus) and Varicella-Zoster Virus (Varicella-Zoster Virus), in which case,ganciclovir can be used as a prodrug to kill transgenic cells in a subject (see, e.g., Clair et al, 1987, animicrob.31:844-849)。

In another embodiment, the gene product of interest can be an AAV protein. In particular, a Rep protein, such as Rep78 or Rep68, or a functional fragment thereof. The nucleotide sequences encoding Rep78 and/or Rep68, if present on the rAAV genome of the invention and expressed in a mammalian cell transduced with a rAAV of the invention, allow integration of the rAAV into the genome of the transduced mammalian cell. Expression of Rep78 and/or Rep68 in rAAV transduced or infected mammalian cells may provide advantages for certain uses of rAAV by allowing long-term or permanent expression of any other gene product of interest introduced by the rAAV in the cell.

in the rAAV vectors of the present invention, at least one nucleotide sequence encoding a gene product of interest for expression in mammalian cells, preferably operably linked to at least one mammalian cell-compatible expression control sequence (e.g., promoter). A number of such promoters are known in the art (see Sambrook and Russel,2001, supra). constitutive promoters that are widely expressed in many cell types, such as CMV promoters, may be used, however, more preferably will be inducible, tissue-specific, cell type-specific or cell cycle-specific promoters, for example, for liver-specific expression, the promoters may be selected from the group consisting of α 1-antitrypsin promoters, thyroid hormone-binding globulin promoters, albumin promoters, LPS (thyroxine-binding globulin) promoters, HCR-apocII hybrid promoters, HCR-hAAT hybrid promoters and E promoters, LP1, HLP, minimal TTR promoters, FVIII promoters, hyperon enhancers, ealb-hAAT. other examples include tumor-selective promoters for tumor-selective expression, and in particular for neural cell tumor cell expression (Pare 2. 2F, 1997, Med. mefa-9, et 5, et 9, et seq. mew.2. for Single cell, Med. sup. 1-2, et al.

AAV is capable of infecting many mammalian cells. See, e.g., Tratschin et al, mol.cell biol.,5(11):3251-3260(1985) and Grimm et al, hum.Gene ther.,10(15):2445-2450 (1999). However, AAV transduction of human synovial fibroblasts is significantly more efficient than in similar murine cells (jenningserial, artritisis, 3:1(2001)), and the cell tropism (tropicity) of AAV varies among serotypes. See, e.g., Davidson et al, proc.natl.acad.sci.usa,97(7):3428-3432(2000) (discussing the differences in mammalian CNS cell tropism and transduction efficiency of AAV2, AAV4, and AAV 5).

As noted, the AAV sequences that can be used in the present invention to produce AAV in insect cells can be derived from the genome of any AAV serotype. In general, AAV serotypes have genomic sequences with significant homology at the amino acid and nucleic acid levels, provide the same set of genetic functions, produce virions that are essentially physically and functionally equivalent, and replicate and assemble by nearly identical mechanisms. For a summary of the genomic sequences and genomic similarities of various AAV serotypes, see, e.g., GenBank Accession number U89790; GenBank Accession number J01901; GenBank accession number AF 043303; GenBank Accession number AF 085716; chlorrini et al (1997, J.Vir.71: 6823-33); srivastava et al (1983, J.Vir.45: 555-64); chlorrini et al (1999, J.Vir.73: 1309-1319); rutledge et al (1998, J.Vir.72: 309-319); and Wu et al (2000, J.Vir.74: 8635-47). Human or simian adeno-associated virus (AAV) serotypes are a preferred source of AAV nucleotide sequences for use in the context of the present invention, more preferably AAV serotypes that normally infect humans (e.g., serotypes 1, 2, 3A, 3B, 4, 5, 6, 7, 8, 9, 10, 11, 12, and 13) or primates (e.g., serotypes 1 and 4).

Preferably, the AAV ITR sequences for use in the context of the present invention are derived from AAV1, AAV2, AAV5 and/or AAV 4. Also, preferably, the Rep52, Rep40, Rep78, and/or Rep68 coding sequences are derived from AAV1, AAV2, and/or AAV 4. The sequences encoding VP1, VP2, and VP3 capsid proteins for use in the context of the present invention may be taken from any known 42 serotypes, more preferably from AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, or AAV9 or newly developed AAV-like particles obtained by, for example, capsid shuffling techniques and AAV capsid libraries. In a preferred embodiment, the sequences encoding VP1, VP2, and VP3 capsid proteins are from AAV5 or AAV8, more preferably from AAV 5.

AAV Rep and ITR sequences are particularly conserved in most serotypes. Rep78 proteins of various AAV serotypes, for example, have over 89% identity, and the total nucleotide sequence identity at the genomic level between AAV2, AAV3A, AAV3B and AAV6 is around 82% (Bantel-Schaal et al, 1999, J.Virol, 73(2): 939-947). In addition, the Rep sequences and ITRs of many AAV serotypes are known to efficiently cross-complement (i.e., functionally replace) corresponding sequences from other serotypes in the production of AAV particles in mammalian cells. US2003148506 reports that AAV Rep and ITR sequences also efficiently cross-complement other AAV Rep and ITR sequences in insect cells.

The AAV VP proteins are known to determine the cellular tropism of AAV virions. The conservation of the VP protein coding sequence in different AAV serotypes is significantly lower than for Rep proteins and genes. The ability of the Rep and ITR sequences to cross-complement the corresponding sequences of other serotypes allows the production of pseudotyped AAV particles comprising the capsid protein of a serotype, such as AAV3, and the Rep and/or ITR sequences of another AAV serotype, such as AAV 2. Such pseudotyped AAV particles are part of the invention.

As noted, modified "AAV" sequences may also be used in the context of the present invention, e.g., for the production of rAAV vectors in insect cells. Such modified sequences include, for example, sequences having at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, or more nucleotide and/or amino acid sequence identity (e.g., sequences having about 75-99% nucleotide sequence identity) to an AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, or AAV9 ITR, Rep, or VP can be used to replace wild-type AAV ITR, Rep, or VP sequences.

Although similar in many respects to other AAV serotypes, AAV5 differs from other human and simian AAV serotypes by a greater degree than other known human and simian serotypes. In this regard, production of AAV5 may differ from production of other serotypes in insect cells. When producing rAAV5 using the methods of the invention, it is preferred to include one or more of the following vectors, in the case of more than one vector, collectively: nucleotide sequences comprising AAV5 ITRs, nucleotide sequences comprising AAV5 Rep52 and/or Rep40 coding sequences, and nucleotide sequences comprising AAV5 Rep78 and/or Rep68 coding sequences. Such ITR and Rep sequences can be desirably modified to obtain efficient production of rAAV5 or pseudotyped rAAV5 vectors in insect cells. For example, the initiation codon of the Rep sequence may be modified.

In a preferred embodiment, the first, second, third and optionally fourth nucleotide sequences are stably integrated in the genome of the insect cell.

A preferred AAV according to the present invention is a virion comprising in its genome at least one nucleotide sequence encoding a gene product of interest, wherein said at least one nucleotide sequence is preferably not a native AAV nucleotide sequence, and wherein said AAV virion comprises a VP1 capsid protein comprising a methionine at amino acid position 1 and a valine at position 2. Even more preferred are AAV virions obtainable from an insect cell as defined above, e.g. in a method as defined below.

An advantage of the AAV virions of the invention is their improved infectivity. Without wishing to be bound by any theory, it appears that infectivity increases as the amount of VP1 protein in the capsid is increased relative to the amount of VP2 and/or VP3 in the capsid in combination with valine at position 2 of VP 1. Infectivity of an AAV virion is herein understood to refer to the transduction efficiency of the transgene comprised in the virion, which can be inferred from the expression rate of the transgene and from the amount or activity of the product expressed by the transgene.

Preferably, the AAV virions of the invention comprise a gene product of interest encoding a polypeptide gene product selected from the group consisting of: CFTR, factor IX, lipoprotein lipase (LPL, preferably LPL S447X; see WO 01/00220), apolipoprotein A1, uridine diphosphate glucuronyl transferase (UGT), retinitis pigmentosa GTPase regulatory interaction protein (RP-GRIP), cytokines or interleukins such as IL-10, dystrophin, PBGD, NaGLU, Treg167, Treg289, EPO, IGF, IFN, GDNF, FOXP3, factor VIII, VEGF, AGXT and insulin. More preferably, the gene product of interest encodes a factor IX or factor VIII protein.

In another aspect, the invention thus relates to a method for producing AAV in an insect cell. Preferably, the method comprises the steps of: (a) culturing an insect cell as defined herein above under conditions such that AAV is produced; and optionally, (b) recovering the AAV. Growth conditions for insect cells in culture and production of heterologous products in insect cells in culture are well known in the art and are described, for example, in the references cited above for molecular engineering of insect cells.

Preferably, the method further comprises the step of affinity purifying the AAV using an anti-AAV antibody (preferably an immobilized antibody). The anti-AAV antibody is preferably a monoclonal antibody. Particularly suitable antibodies are single chain camelidae antibodies or fragments thereof, e.g. as obtainable from camels or llamas (see e.g. muydermans, 2001, Biotechnol).74:277-302). The antibodies used for affinity purification of AAV are preferably antibodies that specifically bind to epitopes on the AAV capsid protein, wherein the epitopes are preferably epitopes present on capsid proteins of more than one AAV serotype. For example, the antibodies may be generated or selected based on specific binding to the AAV2 capsid, but may also be based on specific binding to the AAV1, AAV3, and AAV5 capsids.

In another aspect of the invention, there is provided a method for providing a nucleic acid construct encoding a parvovirus capsid protein, said nucleic acid construct having one or more improved properties, the method comprising:

a) providing a plurality of nucleic acid constructs, each construct comprising:

a nucleotide sequence encoding a parvoviral capsid protein operably linked to an expression control sequence, and at least one parvoviral Inverted Terminal Repeat (ITR) sequence flanking the nucleotide sequence encoding a parvoviral capsid protein operably linked to an expression control sequence;

b) transferring a plurality of nucleic acid constructs into an insect cell capable of expressing a parvoviral Rep protein;

c) placing said insect cell under conditions that allow expression of the parvoviral capsid protein and the parvoviral rep protein such that said nucleic acid construct can be packaged in a parvoviral capsid to provide a parvoviral virion;

e) contacting the parvoviral virion with a target cell to allow infection of the target cell;

f) recovering or identifying the nucleic acid construct from the target cell.

As described in the examples section and above, this method is particularly suitable for first selecting a nucleic acid construct that is highly functional in insect cells with respect to: the constructs are capable of producing large numbers of capsids containing the vector genome, and are also capable of producing capsids containing constructs that are highly efficient in transferring their DNA and subsequently expressing it into target cells.

It is understood that with respect to a plurality of nucleic acid constructs, it is meant constructs having variations with respect to expression control sequences and/or nucleic acid sequences encoding the amino acid sequence of the capsid protein and/or the ITR sequence. Accordingly, any variations thereof are contemplated. For improvement of any property, these may be related to a reference sequence, e.g., a prior art wild-type sequence or nucleic acid construct used to produce AAV capsids in insect cells. Any property that may require improvement in relation to sequences that may be altered in multiple nucleic acid constructs may be considered. Such properties may include, but are not limited to, for example, improved titer, improved yield, improved target cell selectivity.

Creating molecular diversity or mutations is the first step in the methods of the invention. By introducing random point mutations in a reference sequence for which improvement is sought, for example by error-prone (EP) PCR, a plurality of nucleic acids encoding the mutated sequence (i.e. a library of mutated nucleic acids). As mentioned, the random mutation may be included in the non-coding sequence and/or the coding sequence. The frequency of mutations that can be introduced can be varied by varying the amount of template and PCR cycles, as well as the mutagenic primers used. It is understood that when reference is made to a plurality, this relates to 100 or more, preferably 1000 or more, 10000 or more, 100000 or more, or 10000000 or more different sequences, depending on the variation to be introduced in the plurality of nucleic acid constructs. It is understood that the term "library" or "plurality" may have the same meaning herein in so far as they refer to a large number of different sequences, which may for example be related, i.e. have substantial sequence identity. Each member of the library, i.e., each different sequence, may be represented more than 1 time in the library. For example, when a library contains 1000 unique sequences, the library may contain a total of 1000000 sequences. This means that on average 1000 copies are present in the library per library member.

Mutagenesis may be performed in any manner known to those skilled in the art. For example, such mutagenesis can be random, although such mutagenesis can be targeted (i.e., targeted to a specific sequence/structure within a nucleic acid construct, for example). Random mutagenesis can be performed to achieve a low mutation rate, for example, to provide a sequence encoding a Cap protein with 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 or more amino acid changes (as compared to the starting sequence on which the mutagenesis is performed).

Techniques that can be used to perform random mutagenesis include E.coli XL1red, UV irradiation, chemical methods (e.g.deamination, alkylation or base analogue mutagens) or PCR methods (e.g.DNA shuffling, site-directed random mutagenesis or error-prone PCR).

Error-prone PCR is a modification of the standard PCR protocol aimed at altering and enhancing the natural error rate of the polymerase. Taq polymerase can be used because of its naturally high error rate, which error favors the change from AT to GC. However, it is also possible to use alternative forms of polymerase, which bias allows for increased variation in mutation type (i.e. more GC to AT variation).

Error-prone PCR reactions typically contain higher concentrations of MgCl than the underlying PCR reaction₂To stabilize the non-complementary pair. Can also be addedAdding MgCl₂To increase the error rate. Other ways of adjusting the mutation rate include changing the ratio of nucleotides in the reaction, or including nucleotide analogs such as 8-oxo-GTP or dITP. The mutation rate can also be adjusted by changing the effective doubling number by increasing/decreasing the number of cycles or by changing the starting template concentration.

In any case, regardless of the manner in which the mutations are introduced, the resulting plurality of sequences is then cloned into a nucleic acid construct to obtain a plurality of nucleic acid constructs. The nucleic acid construct comprises one or more parvoviral or AAV ITRs flanked by nucleotide sequences encoding parvoviral capsid proteins operably linked to expression control sequences (typically flanked by two AAV ITRs). The nucleic acid construct may also comprise, for example, between the ITRs, an optional reporter gene expression cassette (such as a Green Fluorescent Protein (GFP) expression cassette) under the control of a promoter (such as the CMV and baculovirus p10 promoters). Multiple constructs can then be introduced into a vector of interest, such as a baculovirus vector, to obtain a library of baculoviruses. This can be readily achieved by using common biomolecular techniques (e.g. homologous recombination) and also by using commercially available systems (e.g. Bac-to-Bac). Each baculovirus in the library contains a single nucleic acid construct, wherein the single nucleic acid construct has the expected sequence variation. When generating a baculovirus library, the complexity of the library is preferably maintained (i.e., the amount of unique sequences in the baculovirus library remains approximately the same as compared to the nucleic acid library). Thus, preferably, the nucleic acid construct as defined in step a) of the above method is comprised in a baculovirus vector.

Subsequently, multiple constructs were transferred into insect cells. Preferably, a plurality of baculoviruses is used. This is because with baculovirus, multiplicity of infection can be well controlled. Therefore, when a baculovirus library is used, the multiplicity of infection is preferably kept at 1 or less, preferably at 0.5 or less, more preferably at 0.1 or less. For example, at moi of 0.5, most insect cells will have a single baculovirus/cell, however, a significant fraction of these cells will have two baculoviruses/cells from the library and most cells will not be infected. The number of baculoviruses per cell is governed by the poisson distribution. Reducing moi even further reduces the number of cells with more than 1 baculovirus. However, according to the present invention, it may not be necessary to know the multiplicity of infection. For example, as shown in the examples section, serial dilutions of multiple baculoviruses can also be used, and dilutions can be selected that provide optimal AAV vector library production (e.g., highest titer and/or minimal cross-packaging).

The insect cells to which the various constructs are provided are also capable of expressing parvoviral Rep proteins. For example, additional baculoviruses containing the Rep expression constructs can be used to transfer the Rep expression constructs into cells. Preferably, a relatively high multiplicity of infection is used, such that Rep is not a limiting factor, i.e. when a cell is provided with one of the constructs, the cell will also have a high chance of having a Rep expression construct. Alternatively, a stable cell line containing the Rep expression constructs can be used, which can constitutively express the Rep proteins, or which can inducibly express Rep when one of the constructs is transferred to the cell. In any case, the insect cell capable of expressing parvoviral Rep proteins and provided with one (or more) of the plurality of constructs according to the invention is then subjected to conditions that allow expression of parvoviral capsid proteins and parvoviral Rep proteins such that the nucleic acid construct can be packaged into a parvoviral capsid to provide a parvoviral virion. Typically this involves culturing the cells for a period of time, when for example a baculovirus system is used. Preferably, when using a baculovirus vector system, conditions are selected that do not allow the baculovirus to spread to such an extent that many, if not most, cells will contain several members of the construct library. Preferably, such conditions are chosen such that most cells containing constructs from the library will comprise a single construct and will only produce parvoviral capsids encoded by said construct, which also contain said single construct. When conditions are to be selected in which more than one construct will be contained in an insect cell, where one of the constructs produces infectious or titre AAV, constructs with little or no infectivity will be cross-packaged, making it difficult to determine which construct of all packaged constructs is capable of producing the titre AAV. In other words, having a low cross-wrap allows for a more stringent and more efficient selection.

Subsequently, parvoviral virions are recovered from the insect cells and/or insect cell supernatants. Many methods for recovering parvoviral virions are available and include the methods described in the examples section. In addition, conventional methods such as density (step) gradient centrifugation (iodixanol, CsCl) and/or tangential flow filtration may also be used. Such conventional methods may be useful when, for example, variations are introduced in the capsid sequence that can affect affinity chromatography. Nevertheless, it is also possible to consider affinity chromatography steps comprising specificity, as one of the features based on which the construct can be selected. Thus, improved specific affinity chromatography characteristics may also be one of the characteristics that may be expected to improve. Nevertheless, efficient production and infectivity or titer in insect cells remains a feature that needs to be maintained and/or can be improved.

In another embodiment, a library of parvoviral virions produced by the above method is provided. In yet another embodiment, a parvoviral library comprising various parvoviral vectors is provided, said parvoviral vector library comprising parvoviral vector capsids, wherein each parvoviral capsid contains a parvoviral vector genome comprising an expression cassette for expression of a parvoviral capsid protein. Preferably, the parvoviral vector library comprising various parvoviral vectors comprises parvoviral vector capsids, wherein each parvoviral capsid comprises a parvoviral vector genome comprising an expression cassette for expression of parvoviral capsid proteins in insect cells. More preferably, said parvoviral vector library comprising various parvoviral vectors comprises parvoviral vector capsids, wherein substantially each parvoviral capsid is contained in the library, the parvoviral capsid comprising a parvoviral vector genome comprising an expression cassette for expression of a parvoviral capsid protein (encapsidated therein) in insect cells. Alternatively, as described above, the vector genome may not necessarily contain an expression cassette, but may also contain a sequence identifier by which the parvoviral amino acid sequence (and/or the expression cassette encoding it) in which the vector genome is encapsidated can be identified (see fig. 6). In other words, the library contains substantially parvoviral capsids, which can contain any sequence within which the vector genome is encapsidated, as long as the corresponding parvoviral capsids (i.e., the amino acid sequences thereof) in which it is contained can be identified from the sequences contained within the vector genome.

The length of the specific identifier sequence (see FIG. 6B) which may be considered is preferably at least 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 or 20 nucleotides. Specific identifier sequences which can be considered can be up to 50, 60, 70, 80 or 90 nucleotides in length. When the identifier sequence is at least 15 nucleotides, about 10e9 possible unique combinations are possible. Having a longer sequence identifier may allow for more redundancy and more reliable authentication. It is understood that the sequence identifier may be a priori coupled to a specific capsid sequence. Thus, in this case, when the sequence identifiers are sequenced or detected, the corresponding capsid expression cassettes can be identified by reference to the table. Alternatively, the sequence identifier may be used by capturing and/or sequencing a vector carrier genome (e.g., a baculovirus genome), such that the capsid expression cassette sequence, or portion thereof, associated with the sequence identifier can be determined. Such analysis and/or sequence determination may be performed at a later time. Such means and methods for sequence determination using high throughput sequencing techniques to identify sequences from complex libraries are well known in the art.

The library according to the invention as described above or the library generated as described above may be provided as a crude lysate or as a purified product. In particular, such a library may preferably be generated from a viral vector containing a vector genome and encoding parvoviral capsid proteins. A preferred vector for generating the library may be a baculovirus vector containing an expression cassette for the parvoviral capsid protein active in insect cells. Alternatively, it is easily envisaged that any suitable viral vector library and cell line may be considered for selection, for example an adenovirus, HSV, lentiviral vector based system may be used instead of baculovirus, wherein the expression cassette for the capsid protein is suitable (or selected accordingly) for expression in mammalian cells, for example HeLa cells, 293 cells, CHO cells, a549, 293T, COS. Such alternative carrier vehicles and suitable Cell lines that may be considered are well known in the art, as described, for example, in Gene and Cell Therapy-Therapeutic Mechanisms and systems, 2015, CRC Press, 4 th edition, written by Nancy Smith Templeton. Thus, in alternative embodiments, instead of using baculovirus vectors and insect cells, the means and methods described herein can be readily used for mammalian cells, in combination with a suitable mammalian viral vehicle. In any case, since parvoviral vector libraries provided according to the invention (e.g., AAV vector libraries) are generated using vector vehicles that allow control of copy number per producer cell, the quality of the vector libraries is significantly improved compared to libraries produced from plasmids that do not allow such control.

Subsequently, parvoviral virions are recovered or otherwise partitioned to provide a library of parvoviral vectors, which can be crude lysate or purified products, which are then contacted with selected target cells to allow parvovirus infection of the target cells. Suitable target cells can be selected, which can be suitable target cells for which gene therapy is being developed, such as hepatocytes, kidney cells, neurons. Suitable target cells may be cell lines, such as HeLa cells, HEK293 cells or HuH cells, but also primary cells. It is even envisaged that this includes delivery to a suitable animal model, e.g. rat, mouse, monkey, and may also include various routes of delivery, e.g. intravenous or intramuscular injection, and that the subsequent target cells are candidate organs of choice in such an animal model. In any case, any cell type can be selected and the parvoviral virion contacted therewith in any manner (i.e., in vivo or in vitro) to allow infection, i.e., transfer of the nucleic acid construct contained within the capsid virion to the cell. It is understood that the cells may also be co-infected with adenovirus to assist in the transduction process, e.g. to induce transduction. This may be helpful when, for example, a reporter construct is contained in the nucleic acid construct and it is desired that the selected cell not only allows for efficient transfer of DNA, but also efficient transport within the cell to deliver the nucleic acid construct to the nucleus (see fig. 6). Without being limited by theory, when the capsid sequence is mutated and/or the stoichiometric ratio of VP1, VP2, and VP3 is altered, this can lead to a hindrance of internal transport. For example, a capsid lacking VP1 may infect a cell, but will not enter the nucleus. The capsid containing the nucleic acid construct is then retained in the endosome and targeted for proteolysis by the proteasome. Therefore, it may be of interest to include a selection step for the purpose of the selection process, i.e. in order to achieve efficient delivery of the nucleic acid construct to the nucleus of the cell to allow expression from the nucleic acid construct. This may be, for example, by a reporter gene or any other gene of interest. This may also be a HeLa RC32 cell or the like, wherein the virosomes that achieve efficient delivery of the vector genome are amplified.

Finally, when the cell has been allowed to infect the target cell, preferably allowing efficient transduction, the nucleic acid construct is recovered from the target cell. The nucleic acid construct can be recovered from a whole cell population. The nucleic acid construct may also be recovered from a subset of the population of cells, such as a subset that exhibits reporter transgene expression and is therefore efficiently transduced to the target cell. The nucleic acid construct may also be recovered from a population of whole cells but in particular from nuclei from whole cells. In this way, a nucleic acid construct (and also its accompanying encoded capsid) can be selected that is expected to be good at transducing target cells. The recovered nucleic acid construct may then be sequenced to identify the nucleic acid construct. As noted, the nucleic acid construct may contain an identifier sequence to identify the construct. It is also understood that when, for example, baculovirus vector systems and insect cells have been used for parvovirus vector library generation, and the parvovirus vector genome contains an expression cassette for containing the parvovirus capsid therein, the expression cassette or a portion thereof can be considered an identifier sequence. When the expression cassette has an insect cell promoter and a promoter that is not active in mammalian cells, the expression cassette may not produce an AAV capsid when introduced into mammalian cells. In particular, the part of the nucleic acid construct (or the corresponding identifier sequence) in which the variation is introduced may be sequenced, for example after a PCR reaction, wherein a small part is simply amplified. The entire nucleic acid construct or the entire capsid encoding sequence may also be sequenced. It is understood that sequencing includes high-throughput sequencing or any other suitable sequence method known in the art.

Of particular interest may be the identification of improved sequences. When conditions are chosen such that they are highly limiting, all recovered nucleic acid constructs and their sequences are improved nucleic acid constructs. Thus, recovery of the nucleic acid construct involves improved selection of the nucleic acid construct. Nevertheless, improved sequences derived from recovered nucleic acid constructs can be identified or identified by comparing a population of recovered sequences to, for example, a population of sequences contained in the initially constructed library. A high preponderance of recovered sequences in the recovered population when compared to the initial population is indicative of a desired improved nucleic acid construct. Thus, in addition to recovery of the nucleic acid construct, an additional step may include identifying from the library a nucleic acid construct corresponding to the improved nucleic acid construct. Such identification may include comparison to population sequences from one or more of, for example, an initial library, a baculovirus library containing nucleic acid constructs, a population of nucleic acid constructs contained in a parvovirus capsid.

Once the nucleic acid construct having the improved properties selected for it is provided or identified, the next step is step g) to generate a nucleic acid construct for use in the production of a gene therapy vector comprising a nucleotide sequence encoding a parvoviral capsid protein operably linked to an expression control sequence as recovered in step f). The nucleic acid construct used to generate the gene therapy vector does not have an expression construct of parvoviral capsid proteins flanked by parvoviral ITR sequences. Thus, the nucleic acid construct for producing the gene therapy vector preferably comprises an expression construct for the parvoviral capsid protein, and may optionally further comprise parvoviral components, such as, for example, gene therapy constructs, i.e., therapeutic genes flanked by parvoviral ITRs, and/or Rep expression constructs, all of which are constructed for production compatibility with insect cells. Thus, preferably, the generated nucleic acid construct is comprised in a baculovirus vector or an insect cell. Since AAV viral vectors are good candidates for gene therapy, in particular the parvoviral capsid proteins, parvoviral Rep proteins and/or ITR nucleotide sequences are preferably derived from adeno-associated viruses. It is understood that the recovered nucleic acid construct used to generate the nucleic acid construct for use in generating the gene therapy vector may be the actual physical nucleic acid, e.g., obtained by excising the sequence of interest from the recovered nucleic acid construct. Alternatively, the sequence of interest, e.g. a parvoviral capsid expression cassette or part thereof, can be amplified by a PCR reaction and subsequently used. Likewise, the sequence may be determined and the sequence of interest may be generated de novo (e.g., by a DNA synthesizer).

Since the entire selection process is to identify improved constructs for insect cell-based manufacture of gene therapy vectors for use in medical treatment, in yet another embodiment, a method for producing parvoviral vectors is provided, said method comprising steps a) -g) as described above, wherein insect cells are provided with:

the nucleic acid construct generated for the production of a gene therapy vector

-a nucleic acid construct comprising a nucleotide sequence comprising at least one Inverted Terminal Repeat (ITR) nucleotide sequence; and

-a nucleic acid construct encoding parvoviral Rep proteins capable of expressing parvoviral Rep proteins in insect cells;

wherein the insect cell is cultured under conditions such that a parvoviral vector is produced; and optionally, (b) recovering the produced parvoviral vector. Preferably, the parvoviral vector is an AAV vector. Thus, any of the methods described above for producing an AAV vector having VP1, VP2 and VP3 expression constructs with an out-of-frame start codon preceding the VP1ATG codon are equally applicable to any of the identified improved constructs and resulting nucleic acid constructs for use in producing gene therapy vectors.

In this document and in the claims hereof, the verb "to comprise" and its conjugations is used in its non-limiting sense to mean that items following the word are included, but items not specifically mentioned are not excluded. In addition, reference to an element by the indefinite article "a" or "an" does not exclude the possibility that more than one of the element is present, unless the context clearly requires that there be one and only one of the elements. The indefinite article "a" or "an" thus usually means "at least one".

All patents and references cited in this specification are incorporated herein by reference in their entirety.

The following examples are for illustrative purposes only and are not intended to limit the scope of the present invention in any way.

Examples

1. Introduction to the design reside in

The expression of AAV capsids in the baculovirus expression system (BEVS) requires modification of the expression cassette to facilitate single mRNA transcript leading to the production of the three viral capsid proteins in the correct ratio. Work done by Urabe et al (2002; supra) showed that altering the start codon binding removes one intron splice site, resulting in expression of all three VP proteins in insect cells. Further work suggests that CTG and GTG can be used as efficient start codons for AAV production in the BEVS system. Concomitantly, the alanine at the second position, for example, by introducing it into the AAV5 capsid sequence, results in an AAV5 capsid having a native VP1 to VP3 capsid protein ratio.

However, in a rational design process, limited constructs and combinatorial subsets are possible due to the labor intensive work of generating recombinant baculoviruses. Thus, a series of selective initiation codons (total of 17) were designed using library approach design in conjunction with random background sequences within the AAV5 capsid to determine if there is still room to select for improvement in AAV capsid quality and yield from the BEVS system (see figure 1 for a summary of the approach). The results and methods described below are not limited to AAV5, but are equally applicable to other serotypes and other parvoviruses, and are equally applicable to the selection of improvements in other characteristics of parvoviral gene therapy vectors.

Materials and methods

Construct design and plasmid library

The following alternative putative initiation codons spanning different eukaryotes and prokaryotes were found in the literature and used as possible initiation codons for AAV 5VP1 production: ATT, ATG, ATA, AGA, AGG, AAA, CTG, CTT, CTC, CTA, CGA, CGC, TTG, TGA, TAA, TAG, and GTG. The constructs have the following background design: NNN NNN NNN GNN NNN (SEQ ID NO: 71). WhereinNNNIndicates the insertion of any of the above-described start codons for VP1, and N indicates A, T, C or G, which are randomly and uniformly distributed. The "G" in the first triplet after the start codon is fixed. The theoretical complexity of this library was calculated to be 7.1 × 10⁷(4¹¹x17), the maximum number of unique sequences that can be generated. A subfamily of the initiator codon was synthesized in GeneArt (ThermoFisher) and the entire sequence with AAV 5-encoding capsid sequences and gene expression sequences was cloned into ITR-containing plasmids such that the resulting AAV capsids encapsidate the capsid-encoding genes as transgenes within themselves. Plasmid libraries were generated in GeneArt, where 100 single colonies from the library were Sanger sequenced to confirm complexity and diversity within the library.

Baculovirus libraries

To develop the functionality of the BEVS system and thus screen new designs for its compatibility with the BEVS system, we generated a recombinant baculovirus library from the plasmid library provided above. The theoretical diversity of the library was 7.1x10⁷. In a standard recombination protocol, we used 1. mu.g of donor plasmid (8.12X 10)¹⁰Plasmid molecule) and 1. mu.g of Bsu36I digested BacAMT5 baculovirus scaffold (7.34X 10)⁹A molecule). In the case of a recombination efficiency of 100%, the limiting factor is a baculovirus scaffold representing more than 103 times the theoretical library complexity. The pooled P0 library was amplified in SF9 cells, where baculovirus amplification was expected to be about 1000-fold, resulting in a P1 library representing the complete complex library.

Generation of AAV libraries

For production of AAV libraries, SF9 cells were seeded at 100 ten thousand cells/ml. The MOI (multiplicity of infection) is calculated as follows: MOI is 0.7x virus volume x titer/cell density x cell volume. We determined that the titer of P1 passage of the baculovirus library was approximately 2X10¹¹gc/ml. Mean TCID50 values for baculoviruses were estimated to be about 2 log lower than genomic copy titers. Resulting in an estimated TCID50 value of 2x10⁹And/ml. For the P1 baculovirus library, the calculated infection titer 2x10 was used⁹The first AAV library was generated (MOI 0.5). By inoculating 3L insect cells at 100 ten thousand cells/ml we had an MOI of 0.5 for the capsid/transgene. In other words, there is less than one infectious particle per cell. Since the capsid is also transgenic, (and thus for the capsid) the cassette will be amplified approximately 1000-fold per cell by the replicase. This double infection was also statistically more efficient in terms of poisson distribution than the triple infection. Three additional AAV libraries were generated using estimated MOIs of 5, 25, and 50. AAV libraries generated at an MOI of 0.5 were found to perform best in the selection method.

Purification and quantification of AAV

The material of the AAV library was purified from 3L CLB by 5mLAVB agarose column (affinity chromatography) on Akta Explorer. DNA was isolated and qPCR was performed on each fraction using primers that amplify AAV vector genomic sequences. Thus, we pooled the fractions for a modified TCID50 assay on HeLa RC32 cells to apply selective pressure on the new mutants in the library. See below. The other three AAV productions (

MOI

5, 25 and 50, respectively) were isolated in a similar manner. DNA was isolated from all isolated AAV libraries for Next Generation Sequencing (NGS).

Selective pressure on AAV libraries

A modified TCID50 assay on HeLa RC32 cells (Tessier J, et al.J.Virol.75(1):375-383,2001) was used to select the AAV variants that showed the highest titers. HeLa RC32 cells contained AAV2 replicase and capsid genes integrated into the genome. Following transduction with AAV, the transgene is amplified by replicase and packaged in AAV capsids, which are also produced in HeLa cells. The advantage of this cell line is in principle that the replicase acts as an amplifier for any AAV DNA that enters the nucleus. By limited serial dilution of AAV and infection of HeLa cells, we can selectively amplify only those AAV that successfully completed reaching the nucleus. In other words, AAV capsids and constructs containing/encoding a good ratio of VP1: VP2: VP3 were selected. Serial dilutions for transducing HeLa cells were: 6400 gc/cell, 3200 gc/cell, 1600 gc/cell, 800 gc/cell, 400 gc/cell, and 200 gc/cell.

Isolation of AAV DNA

Two days after transduction, HeLa cells were lysed and DNA isolation was performed to recover AAV vector genomes, where the vector genome reaching the nucleus was expanded in HeLa cells. The isolated DNA was subjected to end-point PCR using a universal primer set for the capsid library prior to submission to Next Generation Sequencing (NGS).

NGS sequencing of multiple libraries

NGS sequencing was performed on isolated DNA from plasmid libraries, P1 passages of baculovirus libraries, production of AAV libraries, and DNA isolated from pooled dilutions transduced from each AAV library. The DNA prepared for each sequencing reaction was sent to baseclean for amplification and barcode technology.

Results

AAV libraries were generated from 0.5MOI infection. After production of the AAV library, the library was used to infect HeLa RC32 cells. Plasmid libraries, baculovirus libraries, AAV libraries and infected HeLa RC32 were processed and analyzed for next generation gene sequencing to determine their complexity. Unique sequences were identified at each step and the copy number of each unique sequence was also determined, the total number of sequences was determined, and the relative percentage for each start codon was determined and plotted (see FIGS. 2A-E). The generated baculovirus library represents an estimated 74% of the complexity of the plasmid library. There was no striking observation between the plasmid library and the baculovirus library regarding the prevalence of the initiation codon (see fig. 2A and 2B), which was expected because no selection pressure was applied thereto. However, when ATG is used as the start codon, this sequence is found minimally in AAV capsids (see, e.g., fig. 2C). Here, ATG represents less than 0.5% of the total library. This low percentage is expected because the strong start codon for VP1 produces mainly VP1 protein, while little or no VP2 and VP3 protein are produced, with VP3 generally being necessary for capsid production. For the rest, no dramatic observations were made regarding the percentage of codon usage between plasmid libraries, baculovirus libraries and AAV capsid libraries, as they were all within the normal range of variation (ranging from about 4-5% to about 8-9%). Finally, AAV libraries typically represent about 96% of the complexity of baculovirus libraries, suggesting a comprehensive shift in the complexity of generating AAV from baculovirus libraries. Finally, when HeLa RC32 cells were infected with the AAV library at a limited series of dilutions, we found that CTG and GTG were the two most abundant start codons in the baculovirus expression system that produced potent AAV viral capsid particles. CTG and GTG together make up almost 50% of all sequences that successfully transduce and infect cells (i.e. vector genomic DNA transferred to the nucleus to allow amplification by HeLa RC32 cells). Strikingly, although only the codon immediately following the start codon was limited to G, mainly the codon following the start codon was found to encode alanine (not shown), confirming that the triplet encoding Ala may be preferred as the second codon in insect cells for expression of VP1, due to the amino acid sequence and/or due to the DNA/RNA sequence. Strikingly, sequences recovered from cells indicate that the AAV library is only 5% complex to recover. This indicates that the selection pressure is significant.

Interestingly, in DNA isolated from HeLa RC32 cells, ATG as the initiation codon was the third highest initiation codon, accounting for approximately 8% of the complete library. This is in contrast to representing only 0.5% in AAV libraries. The first thirty sequences with the VP1 start codon are listed in table 1 below, with the most prevalent sequence that accounts for the majority of the population listed at the top (SEQ id No. 1). Each sequence itself allows for efficient production of AAV capsids when used as a replacement sequence for the context of the VP1 start codon sequence. While each sequence may have some inherent properties per se that allow for efficient production of AAV capsids, in addition, the basic features may be identified from the sequences listed below, which may describe some general rules governing efficient production of a potent AAV from the ATG start codon (i.e., see fig. 5). This may include, but is not necessarily limited to, a (out of frame) start codon preceding the VP1 start codon and/or a GT sequence immediately following the ATG codon, resulting in a valine preferably at position 2 of the VP1 capsid. For most of the 30 clones, an upstream out-of-frame start codon was observed that can serve as a translation initiation site (ATG, CTG, ACG, TTG, and GTG). Such an out-of-frame start codon is expected to result, when translated, in a short peptide with a stop codon following the start codon of VP 1. Likewise, out-of-frame CTT or CTC non-canonical start codons can be identified. Although CTT and CTC are not considered strong non-canonical initiation codons, we observed that a diverse capsid was isolated from HeLa cells that clearly contained both initiation codons. Without being limited by theory, this suggests that an out-of-frame start codon preceding VP1ATG can serve as a decoy translation initiation background for the ribosome, interfering with VP1 translation and allowing pseudoleaky ribosome scanning as can be observed with wild-type AAV. More specifically, the synthesis of (short) peptides from these alternative start codons can allow the ribosome to continue scanning for mRNA transcripts or to restart them. This delayed and leaky initiation may allow translation of VP2 and VP3 from one polycistronic mRNA transcript. Furthermore, this can be demonstrated to be similar to what happens when CTG, GTG, TTG and ACG are introduced as non-canonical start codons (granted European patent No 1945779B 1; granted U.S. patent No 8163543; Urabe et al 2002; supra), allowing ribosomes to regularly not initiate translation at the non-canonical VP1 start codon, which allows for adequate initiation of translation of VP2 and VP3 in a single mRNA transcript from their respective start codons.

Table 1 the first 30 sequences from ATG-containing clones recovered from HeLa RC32 cells.

To confirm that the selection process of the library generated useful new clones, two representative start codon constructs for ATG, CTG, GTG and one representative construct for TAG, TGA, respectively, were selected for recombination into stable baculovirus clones (table 1). These constructs are used to determine the ratio and titer of viral capsid subunits. Furthermore, we wanted to demonstrate that constructs with the ATG initiation codon produce high yields and titers of AAV.

Table 2 unique start codons for baculovirus production and their background sequences.

A unique context for the initiation codon sequence (the underlined VP1 initiation codon) was selected and cloned as a substitution in the AAV5 expression construct sequence (SEQ ID NOS: 70 and 74, where SEQ ID NO:74 corresponds to nts.148-167). SEQ ID NO 31 is the dominant clone selected and identified from the MOI 5 library. Several clones were generated for each candidate and analyzed for VP capsid expression (fig. 3). The initiation codon with its relative background has been successful to varying degrees in generating AAV capsids with good stoichiometry. Notably, in most cases three clones were tested per construct to determine if the baculovirus clones were stable. In this regard, ATG1 has a stable producer (second lane for ATG1 in fig. 3). For ATG2, there were sufficient stable producers, all with good stoichiometric ratios. The CTG1 construct could not be generated, whereas CTG2 produced capsid stoichiometries similar to those described in international patent application WO2015/137802 (data not shown). Similarly, GTG2 also showed good stoichiometry, whereas the amount of TAG (stop codon) produced was very low and TGA (stop codon) resulted in the production of a less capsid of VP 1. Thus, it was surprisingly demonstrated that we were able to generate an efficient AAV capsid construct, AAV5, in which ATG was utilized as the start codon showing a good stoichiometric ratio.

Stable clones for each start codon construct were selected and used to generate AAV carrying the SEAP reporter gene under the control of the CMV promoter. Titers (gc/ml) produced by all AAV constructs were in a similar range. After titration, we transduced Huh7 and HeLa cells at three different MOIs and determined SEAP activity after 48 hours (fig. 5A and 5B). Strikingly, both constructs with the ATG start codon produced capsid titers similar or improved compared to CTG and GTG, whereas capsids lacking VP1(TGA) had no discernible SEAP activity above background as expected. The fact that the predominantly unique clone identified in table 1 encodes valine at position 2 provides supportive evidence that valine can improve potency. These results are consistent with the observations from fig. 3, where these capsids show a VP1: VP2: VP3 stoichiometric ratio very similar to the CTG and GTG constructs.

Sequence listing

<110> Yonike IP Co., Ltd

<120> improved AAV capsid production in insect cells

<130>P6064534PCT

<150>EP 17182429.5

<151>2017-07-20

<160>74

<170>PatentIn version 3.5

<210>1

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 1

<400>1

ctcgatgcat ggtaagcttt 20

<210>2

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 2

<400>2

ctgaatacat ggtcaccttt 20

<210>3

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 3

<400>3

ctaacttaat ggtagcattt 20

<210>4

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 4

<400>4

ctcaatggat ggttagtttt 20

<210>5

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 5

<400>5

ctcgacgtat ggtcacattt 20

<210>6

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 6

<400>6

ctccctgaat ggcattgttt 20

<210>7

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 7

<400>7

ctagcacgat ggcgtcattt 20

<210>8

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 8

<400>8

ctgaccgcat ggcgacgttt 20

<210>9

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 9

<400>9

ctggagatat ggtgagtttt 20

<210>10

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 10

<400>10

cttgttttat ggtaagtttt 20

<210>11

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 11

<400>11

ctcagttgat ggtcagcttt 20

<210>12

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 12

<400>12

ctacttgtat ggtagctttt 20

<210>13

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 13

<400>13

ctcgatgcat ggcaagcttt 20

<210>14

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 14

<400>14

ctgttagaat ggcgacgttt 20

<210>15

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 15

<400>15

ctcgaccaat gggaacgttt 20

<210>16

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 16

<400>16

ctggcgtcat ggggtcgttt 20

<210>17

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 17

<400>17

ctcgatgcat ggtaagctct 20

<210>18

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 18

<400>18

ctcgatgcat ggtgagcttt 20

<210>19

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 19

<400>19

ctcgatgcat ggtaagcctt 20

<210>20

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 20

<400>20

ctcctcggat ggcgtcattt 20

<210>21

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 21

<400>21

cttgggcgat ggtttcattt 20

<210>22

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 22

<400>22

ctaattgaat ggcggagttt 20

<210>23

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 23

<400>23

ctcgatgcat ggtaggcttt 20

<210>24

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 24

<400>24

ctcgatgcat ggtaagcttc 20

<210>25

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 25

<400>25

ctttgcttat ggtaaatttt 20

<210>26

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 26

<400>26

ctcgacgcat ggtaagcttt 20

<210>27

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 27

<400>27

ctcacttgat ggcttaattt 20

<210>28

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 28

<400>28

ctcagggaat gggattcttt 20

<210>29

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 29

<400>29

cttattctat ggtaagtttt 20

<210>30

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence 30

<400>30

ctcggtgcat ggtaagcttt 20

<210>31

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence ATG2

<400>31

ctcgatgcat ggtaagcttt 20

<210>32

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 1

<400>32

cucgaugcau gguaagcuuu 20

<210>33

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 2

<400>33

cugaauacau ggucaccuuu 20

<210>34

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 3

<400>34

cuaacuuaau gguagcauuu 20

<210>35

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 4

<400>35

cucaauggau gguuaguuuu 20

<210>36

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 5

<400>36

cucgacguau ggucacauuu 20

<210>37

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 6

<400>37

cucccugaau ggcauuguuu 20

<210>38

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 7

<400>38

cuagcacgau ggcgucauuu 20

<210>39

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 8

<400>39

cugaccgcau ggcgacguuu 20

<210>40

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 9

<400>40

cuggagauau ggugaguuuu 20

<210>41

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 10

<400>41

cuuguuuuau gguaaguuuu 20

<210>42

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 11

<400>42

cucaguugau ggucagcuuu 20

<210>43

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 12

<400>43

cuacuuguau gguagcuuuu 20

<210>44

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 13

<400>44

cucgaugcau ggcaagcuuu 20

<210>45

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 14

<400>45

cuguuagaau ggcgacguuu 20

<210>46

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 15

<400>46

cucgaccaau gggaacguuu 20

<210>47

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 16

<400>47

cuggcgucau ggggucguuu 20

<210>48

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 17

<400>48

cucgaugcau gguaagcucu 20

<210>49

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 18

<400>49

cucgaugcau ggugagcuuu 20

<210>50

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 19

<400>50

cucgaugcau gguaagccuu 20

<210>51

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 20

<400>51

cuccucggau ggcgucauuu 20

<210>52

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 21

<400>52

cuugggcgau gguuucauuu 20

<210>53

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 22

<400>53

cuaauugaau ggcggaguuu 20

<210>54

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 23

<400>54

cucgaugcau gguaggcuuu 20

<210>55

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 24

<400>55

cucgaugcau gguaagcuuc 20

<210>56

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 25

<400>56

cuuugcuuau gguaaauuuu 20

<210>57

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 26

<400>57

cucgacgcau gguaagcuuu 20

<210>58

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 27

<400>58

cucacuugau ggcuuaauuu 20

<210>59

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 28

<400>59

cucagggaau gggauucuuu 20

<210>60

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 29

<400>60

cuuauucuau gguaaguuuu 20

<210>61

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence 30

<400>61

cucggugcau gguaagcuuu 20

<210>62

<211>20

<212>RNA

<213>Artificial Sequence

<220>

<223>RNA sequence derived from ATG containing sequence ATG2

<400>62

cugucgucau ggugucguuu 20

<210>63

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>CTG1

<400>63

ctcgtgccct ggcttcgttt 20

<210>64

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>CTG2

<400>64

cttgatgtct ggccactttt 20

<210>65

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>CTG1

<400>65

cttccactgt ggcctccttt 20

<210>66

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>CTG2

<400>66

cttccgccgt ggcgtcgttt 20

<210>67

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>TAG1

<400>67

ctgcccccta ggaccgtttt 20

<210>68

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>TGA1

<400>68

cttcaccctg agcgcaattt 20

<210>69

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>Construct

<220>

<221>misc_feature

<222>(3)..(8)

<223>n is a, c, g, or t

<220>

<221>misc_feature

<222>(13)..(17)

<223>n is a, c, g, or t

<400>69

ctnnnnnnat ggnnnnnttt 20

<210>70

<211>2622

<212>DNA

<213>Artificial Sequence

<220>

<223>AAV5 expression construct sequence

<220>

<221>promoter

<222>(1)..(155)

<223>Polyhedrin Promoter

<220>

<221>misc_feature

<222>(105)..(112)

<223>Transcriptional start

<220>

<221>misc_feature

<222>(105)..(2622)

<223>mRNA transcript

<220>

<221>misc_feature

<222>(156)..(158)

<223>Translational startcodon VP1

<220>

<221>misc_feature

<222>(156)..(2333)

<223>CDS ORF: AAV5 Capsid

<220>

<221>polyA_signal

<222>(2382)..(2622)

<223>poly A

<400>70

tgtaatgaga cgcacaaact aatatcacaa actggaaatg tctatcaata tatagttgct 60

gatatcatgg agataattaa aatgataacc atctcgcaaa taaataagta ttttactgtt 120

ttcgtaacag ttttgtaata aaaaaaccta taaatatggt ctcttttgtt gatcacccac 180

ccgattggtt ggaagaagtt ggtgaaggtc ttcgcgagtt tttgggcctt gaagcgggcc 240

caccgaaacc aaaacccaat cagcagcatc aagatcaagc ccgtggtctt gtgctgcctg 300

gttataacta tctcggaccc ggaaacggtc tcgatcgagg agagcctgtc aacagggcag 360

acgaggtcgc gcgagagcac gacatctcgt acaacgagca gcttgaggcg ggagacaacc 420

cctacctcaa gtacaaccac gcggacgccg agtttcagga gaagctcgcc gacgacacat 480

ccttcggggg aaacctcgga aaggcagtct ttcaggccaa gaaaagggtt ctcgaacctt 540

ttggcctggt tgaagagggt gctaagacgg cccctaccgg aaagcggata gacgaccact 600

ttccaaaaag aaagaaggct cggaccgaag aggactccaa gccttccacc tcgtcagacg 660

ccgaagctgg acccagcgga tcccagcagc tgcaaatccc agcccaacca gcctcaagtt 720

tgggagctga tacaatgtct gcgggaggtg gcggcccatt gggcgacaat aaccaaggtg 780

ccgatggagt gggcaatgcc tcgggagatt ggcattgcga ttccacgtgg atgggggaca 840

gagtcgtcac caagtccacc cgaacctggg tgctgcccag ctacaacaac caccagtacc 900

gagagatcaa aagcggctcc gtcgacggaa gcaacgccaa cgcctacttt ggatacagca 960

ccccctgggg gtactttgac tttaaccgct tccacagcca ctggagcccc cgagactggc 1020

aaagactcat caacaactac tggggcttca gaccccggtc cctcagagtc aaaatcttca 1080

acattcaagt caaagaggtc acggtgcagg actccaccac caccatcgcc aacaacctca 1140

cctccaccgt ccaagtgttt acggacgacg actaccagct gccctacgtc gtcggcaacg 1200

ggaccgaggg atgcctgccg gccttccctc cgcaggtctt tacgctgccg cagtacggtt 1260

acgcgacgct gaaccgcgac aacacagaaa atcccaccga gaggagcagc ttcttctgcc 1320

tagagtactt tcccagcaag atgctgagaa cgggcaacaa ctttgagttt acctacaact 1380

ttgaggaggt gcccttccac tccagcttcg ctcccagtca gaacctgttc aagctggcca 1440

acccgctggt ggaccagtac ttgtaccgct tcgtgagcac aaataacact ggcggagtcc 1500

agttcaacaa gaacctggcc gggagatacg ccaacaccta caaaaactgg ttcccggggc 1560

ccatgggccg aacccagggc tggaacctgg gctccggggt caaccgcgcc agtgtcagcg 1620

ccttcgccac gaccaatagg atggagctcg agggcgcgag ttaccaggtg cccccgcagc 1680

cgaacggcat gaccaacaac ctccagggca gcaacaccta tgccctggag aacactatga 1740

tcttcaacag ccagccggcg aacccgggca ccaccgccac gtacctcgag ggcaacatgc 1800

tcatcaccag cgagagcgag acgcagccgg tgaaccgcgt ggcgtacaac gtcggcgggc 1860

agatggccac caacaaccag agctccacca ctgcccccgc gaccggcacg tacaacctcc 1920

aggaaatcgt gcccggcagc gtgtggatgg agagggacgt gtacctccaa ggacccatct 1980

gggccaagat cccagagacg ggggcgcact ttcacccctc tccggccatg ggcggattcg 2040

gactcaaaca cccaccgccc atgatgctca tcaagaacac gcctgtgccc ggaaatatca 2100

ccagcttctc ggacgtgccc gtcagcagct tcatcaccca gtacagcacc gggcaggtca 2160

ccgtggagat ggagtgggag ctcaagaagg aaaactccaa gaggtggaac ccagagatcc 2220

agtacacaaa caactacaac gacccccagt ttgtggactt tgccccggac agcaccgggg 2280

aatacagaac caccagacct atcggaaccc gataccttac ccgacccctt taatctagag 2340

cctgcagtct cgacaagcta gcttgtcgag aagtactaga ggatcataat cagccatacc 2400

acatttgtag aggttttact tgctttaaaa aacctcccac acctccccct gaacctgaaa 2460

cataaaatga atgcaattgt tgttgttaac ttgtttattg cagcttataa tggttacaaa 2520

taaagcaata gcatcacaaa tttcacaaat aaagcatttt tttcactgca ttctagttgt 2580

ggtttgtcca aactcatcaa tgtatcttat catgtctgga tc 2622

<210>71

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>Construct

<220>

<221>misc_feature

<222>(3)..(11)

<223>n is a, c, g, or t

<220>

<221>misc_feature

<222>(9)..(9)

<223>N is part of astart codon and can be any of a, c, t, or g

<220>

<221>misc_feature

<222>(10)..(10)

<223>N is part of a start codon and can be any of t, g, or a

<220>

<221>misc_feature

<222>(11)..(11)

<223>N is part of a start codon and can be any of a, c, t, or g

<220>

<221>misc_feature

<222>(13)..(17)

<223>n is a, c, g, or t

<400>71

ctnnnnnnnn ngnnnnnttt 20

<210>72

<211>6

<212>PRT

<213>Artificial Sequence

<220>

<223>amino acid sequence 1

<400>72

Met His His Gly Lys Leu

1 5

<210>73

<211>4

<212>PRT

<213>Artificial Sequence

<220>

<223>Amino acid sequence 2

<400>73

Met Glu Ile Trp

1

<210>74

<211>20

<212>DNA

<213>Artificial Sequence

<220>

<223>ATG containing sequence derived from SEQ ID NO:70

<400>74

ctataaatat ggtctctttt 20

Claims

1. A nucleic acid construct comprising an expression control sequence for expressing a nucleotide sequence comprising an open reading frame in an insect cell, wherein the open reading frame sequence encodes:

i) adeno-associated virus (AAV) capsid proteins VP1, VP2, and VP 3; and

ii) ATG translation initiation codon for VP 1;

2. The nucleic acid construct according to claim 1, wherein the selector initiation codon is selected from the group consisting of: CTG, ATG, ACG, TTG, GTG, CTC and CTT.

3. A nucleic acid construct according to claim 1 or claim 2, wherein the nucleotide sequence comprises a selective open reading frame, beginning with the selective initiation codon, encompassing the ATG translation initiation codon for VP 1.

4.A nucleic acid construct according to claim 3, wherein said alternative open reading frame following said alternative start codon encodes a peptide of up to 20 amino acids.

5. A nucleic acid construct according to any one of claims 1 to 4, wherein the nucleotide sequence adjacent to the open reading frame and comprising the selector start codon is nucleotide residues 1-8 of SEQ ID No. 1.

6. The nucleic acid construct according to claim 5, wherein the open reading frame comprising the ATG translation start codon for VP1 has the nucleotide sequence of SEQ ID NO 1, wherein residues 9-11 represent the ATG translation start codon for VP 1.

7. The nucleic acid construct according to any one of claims 1 to 6, wherein said second codon of said open reading frame encodes an amino acid residue selected from the group consisting of alanine, glycine, valine, aspartic acid and glutamic acid.

8. A nucleic acid construct according to any one of claims 1 to 7, wherein the AAV capsid protein is an AAV serotype capsid protein.

9. The nucleic acid construct according to any one of claims 1 to 8, wherein the nucleic acid construct comprises a promoter selected from the group consisting of: polyhedral promoter, p10 promoter, 4xHsp27 EcRE + minimum Hsp70 promoter, deltaE1 promoter and E1 promoter.

10. The nucleic acid construct according to any of claims 1 to 9, wherein the nucleic acid construct is a baculovirus vector.

11. An insect cell comprising a nucleic acid construct according to any one of claims 1-10.

12. An insect cell according to claim 11, wherein the insect cell further comprises:

(a) a second nucleotide sequence comprising at least one AAV Inverted Terminal Repeat (ITR) nucleotide sequence;

(b) a third nucleotide sequence comprising a Rep78 or Rep68 coding sequence, said Rep78 or Rep68 coding sequence being operably linked to expression control sequences for expression in an insect cell.

(c) Optionally, a fourth nucleotide sequence comprising a Rep52 or Rep40 coding sequence, said Rep52 or Rep40 coding sequence being operably linked to expression control sequences for expression in an insect cell.

13. A method for producing AAV in an insect cell, the method comprising the steps of: (a) culturing an insect cell as defined in claim 11 or claim 12 under conditions such that AAV is produced; and optionally, (b) recovering the AAV.

14. A method for providing a nucleic acid construct encoding a parvoviral capsid protein, said nucleic acid construct having one or more improved properties, said method comprising:

a) providing a plurality of nucleic acid constructs, each construct comprising:

a parvoviral capsid protein-encoding nucleotide sequence operably linked to an expression control sequence and at least one parvoviral Inverted Terminal Repeat (ITR) sequence flanked by said parvoviral capsid protein-encoding nucleotide sequence operably linked to an expression control sequence;

b) transferring the plurality of nucleic acid constructs into an insect cell capable of expressing parvoviral Rep proteins;

d) recovering parvoviral virions from the insect cells and/or insect cell supernatants.

f) recovering the nucleic acid construct from the target cell.

15. The method according to claim 14, wherein said nucleic acid construct defined in step a) is comprised in a baculovirus vector.

16. A method according to claim 14 or claim 15, further comprising step g): generating a nucleic acid construct for producing a gene therapy vector, said nucleic acid construct comprising the nucleotide sequence encoding a parvoviral capsid protein operably linked to an expression control sequence recovered in step f).